diff --git "a/checkpoint-23375/trainer_state.json" "b/checkpoint-23375/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-23375/trainer_state.json" @@ -0,0 +1,163658 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.23375, + "eval_steps": 500, + "global_step": 23375, + "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 + } + ], + "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": 9.26670151745536e+17, + "train_batch_size": 512, + "trial_name": null, + "trial_params": null +}