| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.995008319467554, | |
| "eval_steps": 500, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0033277870216306157, | |
| "grad_norm": 0.8612529340183872, | |
| "learning_rate": 2.2222222222222225e-06, | |
| "loss": 1.2888, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0066555740432612314, | |
| "grad_norm": 0.8594848070789138, | |
| "learning_rate": 4.444444444444445e-06, | |
| "loss": 1.2889, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009983361064891847, | |
| "grad_norm": 0.8081284978213048, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.2509, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013311148086522463, | |
| "grad_norm": 0.8750160699999223, | |
| "learning_rate": 8.88888888888889e-06, | |
| "loss": 1.3059, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016638935108153077, | |
| "grad_norm": 0.7989695171242613, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 1.2484, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.019966722129783693, | |
| "grad_norm": 0.822931788451026, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 1.2503, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02329450915141431, | |
| "grad_norm": 0.786357778672969, | |
| "learning_rate": 1.5555555555555555e-05, | |
| "loss": 1.248, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026622296173044926, | |
| "grad_norm": 0.7416721544440852, | |
| "learning_rate": 1.777777777777778e-05, | |
| "loss": 1.1911, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.029950083194675542, | |
| "grad_norm": 0.7217074889785909, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1985, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.033277870216306155, | |
| "grad_norm": 0.6101234849945351, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.1292, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.036605657237936774, | |
| "grad_norm": 0.5359417555621394, | |
| "learning_rate": 2.4444444444444445e-05, | |
| "loss": 1.0897, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03993344425956739, | |
| "grad_norm": 0.5029047876713094, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 1.0364, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04326123128119801, | |
| "grad_norm": 0.4852688032790646, | |
| "learning_rate": 2.8888888888888888e-05, | |
| "loss": 1.0105, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04658901830282862, | |
| "grad_norm": 0.552228587853467, | |
| "learning_rate": 3.111111111111111e-05, | |
| "loss": 0.9801, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04991680532445923, | |
| "grad_norm": 0.5455447178121284, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.8954, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05324459234608985, | |
| "grad_norm": 0.5343727527823027, | |
| "learning_rate": 3.555555555555556e-05, | |
| "loss": 0.8542, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.056572379367720464, | |
| "grad_norm": 0.5587140871586306, | |
| "learning_rate": 3.777777777777778e-05, | |
| "loss": 0.8462, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.059900166389351084, | |
| "grad_norm": 0.49399944094203235, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7757, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0632279534109817, | |
| "grad_norm": 0.49687965586156235, | |
| "learning_rate": 4.222222222222222e-05, | |
| "loss": 0.7378, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06655574043261231, | |
| "grad_norm": 0.48181629837625173, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.677, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06988352745424292, | |
| "grad_norm": 0.40171198426571264, | |
| "learning_rate": 4.666666666666667e-05, | |
| "loss": 0.6445, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07321131447587355, | |
| "grad_norm": 0.38662649847652203, | |
| "learning_rate": 4.888888888888889e-05, | |
| "loss": 0.5949, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07653910149750416, | |
| "grad_norm": 0.3142365961791992, | |
| "learning_rate": 5.111111111111111e-05, | |
| "loss": 0.5811, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07986688851913477, | |
| "grad_norm": 0.24897220339598153, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 0.5318, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08319467554076539, | |
| "grad_norm": 0.26655965329076836, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 0.5482, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08652246256239601, | |
| "grad_norm": 0.211807210054923, | |
| "learning_rate": 5.7777777777777776e-05, | |
| "loss": 0.5288, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08985024958402663, | |
| "grad_norm": 0.2140288218888823, | |
| "learning_rate": 6e-05, | |
| "loss": 0.5143, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09317803660565724, | |
| "grad_norm": 0.20744703999300954, | |
| "learning_rate": 6.222222222222222e-05, | |
| "loss": 0.5146, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09650582362728785, | |
| "grad_norm": 0.19035976993790368, | |
| "learning_rate": 6.444444444444446e-05, | |
| "loss": 0.5241, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.09983361064891846, | |
| "grad_norm": 0.20430817161666623, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.4828, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10316139767054909, | |
| "grad_norm": 0.21019189279877437, | |
| "learning_rate": 6.88888888888889e-05, | |
| "loss": 0.489, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1064891846921797, | |
| "grad_norm": 0.21106078406516807, | |
| "learning_rate": 7.111111111111112e-05, | |
| "loss": 0.4828, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10981697171381032, | |
| "grad_norm": 0.17924687301689285, | |
| "learning_rate": 7.333333333333333e-05, | |
| "loss": 0.4872, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11314475873544093, | |
| "grad_norm": 0.1465786115381622, | |
| "learning_rate": 7.555555555555556e-05, | |
| "loss": 0.4638, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11647254575707154, | |
| "grad_norm": 0.13968758145514554, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 0.4799, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.11980033277870217, | |
| "grad_norm": 0.12958125612856886, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4666, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12312811980033278, | |
| "grad_norm": 0.14229119201892915, | |
| "learning_rate": 8.222222222222222e-05, | |
| "loss": 0.4495, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1264559068219634, | |
| "grad_norm": 0.12625913990615895, | |
| "learning_rate": 8.444444444444444e-05, | |
| "loss": 0.4541, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.129783693843594, | |
| "grad_norm": 0.10972410205955695, | |
| "learning_rate": 8.666666666666667e-05, | |
| "loss": 0.4426, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13311148086522462, | |
| "grad_norm": 0.11518014129656727, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 0.4529, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13643926788685523, | |
| "grad_norm": 0.11197657135878636, | |
| "learning_rate": 9.111111111111112e-05, | |
| "loss": 0.4529, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.13976705490848584, | |
| "grad_norm": 0.10824932596781353, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 0.4447, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.14309484193011648, | |
| "grad_norm": 0.10353039827060762, | |
| "learning_rate": 9.555555555555557e-05, | |
| "loss": 0.433, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1464226289517471, | |
| "grad_norm": 0.10315339802260473, | |
| "learning_rate": 9.777777777777778e-05, | |
| "loss": 0.4335, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1497504159733777, | |
| "grad_norm": 0.10889946997126505, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4363, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15307820299500832, | |
| "grad_norm": 0.09771419020249908, | |
| "learning_rate": 0.00010222222222222222, | |
| "loss": 0.4309, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15640599001663893, | |
| "grad_norm": 0.10089274928519984, | |
| "learning_rate": 0.00010444444444444445, | |
| "loss": 0.4224, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.15973377703826955, | |
| "grad_norm": 0.11234151762857073, | |
| "learning_rate": 0.00010666666666666667, | |
| "loss": 0.4322, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16306156405990016, | |
| "grad_norm": 0.10199980209755123, | |
| "learning_rate": 0.00010888888888888889, | |
| "loss": 0.4234, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16638935108153077, | |
| "grad_norm": 0.10752286330191117, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 0.4223, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16971713810316139, | |
| "grad_norm": 0.10291617448028366, | |
| "learning_rate": 0.00011333333333333334, | |
| "loss": 0.4144, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.17304492512479203, | |
| "grad_norm": 0.09999211805467421, | |
| "learning_rate": 0.00011555555555555555, | |
| "loss": 0.4234, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17637271214642264, | |
| "grad_norm": 0.1047516390814671, | |
| "learning_rate": 0.00011777777777777779, | |
| "loss": 0.4283, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.17970049916805325, | |
| "grad_norm": 0.1032089523624031, | |
| "learning_rate": 0.00012, | |
| "loss": 0.4006, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18302828618968386, | |
| "grad_norm": 0.11279217573067286, | |
| "learning_rate": 0.00012222222222222224, | |
| "loss": 0.4208, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18635607321131448, | |
| "grad_norm": 0.10146777851716919, | |
| "learning_rate": 0.00012444444444444444, | |
| "loss": 0.4126, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1896838602329451, | |
| "grad_norm": 0.1020643325129258, | |
| "learning_rate": 0.00012666666666666666, | |
| "loss": 0.4117, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1930116472545757, | |
| "grad_norm": 0.10065970110765685, | |
| "learning_rate": 0.00012888888888888892, | |
| "loss": 0.4072, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.19633943427620631, | |
| "grad_norm": 0.10006107134360871, | |
| "learning_rate": 0.00013111111111111111, | |
| "loss": 0.3963, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.19966722129783693, | |
| "grad_norm": 0.10249298064083277, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.4084, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20299500831946754, | |
| "grad_norm": 0.10190705587589162, | |
| "learning_rate": 0.00013555555555555556, | |
| "loss": 0.3872, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20632279534109818, | |
| "grad_norm": 0.10127645537591388, | |
| "learning_rate": 0.0001377777777777778, | |
| "loss": 0.398, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2096505823627288, | |
| "grad_norm": 0.10066017880500971, | |
| "learning_rate": 0.00014, | |
| "loss": 0.3928, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2129783693843594, | |
| "grad_norm": 0.10626440268866089, | |
| "learning_rate": 0.00014222222222222224, | |
| "loss": 0.4017, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.21630615640599002, | |
| "grad_norm": 0.1024135678046055, | |
| "learning_rate": 0.00014444444444444444, | |
| "loss": 0.396, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.21963394342762063, | |
| "grad_norm": 0.10062558608579443, | |
| "learning_rate": 0.00014666666666666666, | |
| "loss": 0.3906, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.22296173044925124, | |
| "grad_norm": 0.10847597671209852, | |
| "learning_rate": 0.0001488888888888889, | |
| "loss": 0.3941, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22628951747088186, | |
| "grad_norm": 0.10861449014555247, | |
| "learning_rate": 0.0001511111111111111, | |
| "loss": 0.3756, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.22961730449251247, | |
| "grad_norm": 0.10229080418602543, | |
| "learning_rate": 0.00015333333333333334, | |
| "loss": 0.3834, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.23294509151414308, | |
| "grad_norm": 0.11112902567990897, | |
| "learning_rate": 0.00015555555555555556, | |
| "loss": 0.3956, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23627287853577372, | |
| "grad_norm": 0.10153369840434048, | |
| "learning_rate": 0.0001577777777777778, | |
| "loss": 0.3984, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.23960066555740434, | |
| "grad_norm": 0.10242573526196254, | |
| "learning_rate": 0.00016, | |
| "loss": 0.3797, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24292845257903495, | |
| "grad_norm": 0.108304072708511, | |
| "learning_rate": 0.00016222222222222224, | |
| "loss": 0.3941, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24625623960066556, | |
| "grad_norm": 0.11019946559484159, | |
| "learning_rate": 0.00016444444444444444, | |
| "loss": 0.3799, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.24958402662229617, | |
| "grad_norm": 0.10764472264911762, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 0.3922, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2529118136439268, | |
| "grad_norm": 0.10355135311142791, | |
| "learning_rate": 0.00016888888888888889, | |
| "loss": 0.3746, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2562396006655574, | |
| "grad_norm": 0.10562152472006703, | |
| "learning_rate": 0.0001711111111111111, | |
| "loss": 0.3901, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.259567387687188, | |
| "grad_norm": 0.11288891534245112, | |
| "learning_rate": 0.00017333333333333334, | |
| "loss": 0.392, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2628951747088186, | |
| "grad_norm": 0.10003967032313141, | |
| "learning_rate": 0.00017555555555555556, | |
| "loss": 0.3749, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26622296173044924, | |
| "grad_norm": 0.10423327971773556, | |
| "learning_rate": 0.00017777777777777779, | |
| "loss": 0.3855, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26955074875207985, | |
| "grad_norm": 0.11928436009336384, | |
| "learning_rate": 0.00018, | |
| "loss": 0.3791, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.27287853577371046, | |
| "grad_norm": 0.104697634305105, | |
| "learning_rate": 0.00018222222222222224, | |
| "loss": 0.3841, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2762063227953411, | |
| "grad_norm": 0.10515256514605788, | |
| "learning_rate": 0.00018444444444444446, | |
| "loss": 0.38, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2795341098169717, | |
| "grad_norm": 0.10259049055692239, | |
| "learning_rate": 0.0001866666666666667, | |
| "loss": 0.3802, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.28286189683860236, | |
| "grad_norm": 0.11264324890880631, | |
| "learning_rate": 0.00018888888888888888, | |
| "loss": 0.3772, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28618968386023297, | |
| "grad_norm": 0.10729749794235449, | |
| "learning_rate": 0.00019111111111111114, | |
| "loss": 0.3764, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2895174708818636, | |
| "grad_norm": 0.10573289695745294, | |
| "learning_rate": 0.00019333333333333333, | |
| "loss": 0.3718, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2928452579034942, | |
| "grad_norm": 0.10491856929503432, | |
| "learning_rate": 0.00019555555555555556, | |
| "loss": 0.3675, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2961730449251248, | |
| "grad_norm": 0.11063489273632207, | |
| "learning_rate": 0.00019777777777777778, | |
| "loss": 0.3751, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2995008319467554, | |
| "grad_norm": 0.10948207056739194, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3807, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.30282861896838603, | |
| "grad_norm": 0.10778882351835649, | |
| "learning_rate": 0.0001999992478591656, | |
| "loss": 0.3843, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.30615640599001664, | |
| "grad_norm": 0.1046696274689116, | |
| "learning_rate": 0.00019999699144797678, | |
| "loss": 0.375, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.30948419301164726, | |
| "grad_norm": 0.10632509573626704, | |
| "learning_rate": 0.00019999323080037624, | |
| "loss": 0.3732, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.31281198003327787, | |
| "grad_norm": 0.11021641453155047, | |
| "learning_rate": 0.00019998796597293477, | |
| "loss": 0.3877, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3161397670549085, | |
| "grad_norm": 0.11267301018826237, | |
| "learning_rate": 0.00019998119704485014, | |
| "loss": 0.378, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3194675540765391, | |
| "grad_norm": 0.11039050804462736, | |
| "learning_rate": 0.00019997292411794618, | |
| "loss": 0.388, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3227953410981697, | |
| "grad_norm": 0.10218037146823301, | |
| "learning_rate": 0.00019996314731667096, | |
| "loss": 0.3728, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3261231281198003, | |
| "grad_norm": 0.10292285669044796, | |
| "learning_rate": 0.00019995186678809513, | |
| "loss": 0.3693, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.32945091514143093, | |
| "grad_norm": 0.10142763460343682, | |
| "learning_rate": 0.0001999390827019096, | |
| "loss": 0.3752, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.33277870216306155, | |
| "grad_norm": 0.10624594045436814, | |
| "learning_rate": 0.00019992479525042303, | |
| "loss": 0.3808, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33610648918469216, | |
| "grad_norm": 0.10398765071596988, | |
| "learning_rate": 0.00019990900464855892, | |
| "loss": 0.3709, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.33943427620632277, | |
| "grad_norm": 0.09718990856275914, | |
| "learning_rate": 0.0001998917111338525, | |
| "loss": 0.3595, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3427620632279534, | |
| "grad_norm": 0.10173483817840272, | |
| "learning_rate": 0.00019987291496644676, | |
| "loss": 0.3752, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.34608985024958405, | |
| "grad_norm": 0.09816057882008182, | |
| "learning_rate": 0.00019985261642908918, | |
| "loss": 0.3695, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.34941763727121466, | |
| "grad_norm": 0.10448818908807213, | |
| "learning_rate": 0.00019983081582712685, | |
| "loss": 0.3641, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3527454242928453, | |
| "grad_norm": 0.0981262845819303, | |
| "learning_rate": 0.00019980751348850219, | |
| "loss": 0.368, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3560732113144759, | |
| "grad_norm": 0.09970854948850973, | |
| "learning_rate": 0.00019978270976374812, | |
| "loss": 0.3695, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3594009983361065, | |
| "grad_norm": 0.09886665805923836, | |
| "learning_rate": 0.00019975640502598244, | |
| "loss": 0.3771, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3627287853577371, | |
| "grad_norm": 0.102456021944177, | |
| "learning_rate": 0.00019972859967090252, | |
| "loss": 0.3617, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.36605657237936773, | |
| "grad_norm": 0.09449799618945273, | |
| "learning_rate": 0.0001996992941167792, | |
| "loss": 0.3568, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.36938435940099834, | |
| "grad_norm": 0.1101791213546773, | |
| "learning_rate": 0.00019966848880445062, | |
| "loss": 0.3796, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.37271214642262895, | |
| "grad_norm": 0.09819196255909177, | |
| "learning_rate": 0.00019963618419731538, | |
| "loss": 0.3574, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.37603993344425957, | |
| "grad_norm": 0.09919515660596091, | |
| "learning_rate": 0.00019960238078132577, | |
| "loss": 0.3675, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3793677204658902, | |
| "grad_norm": 0.09444247105680212, | |
| "learning_rate": 0.00019956707906498044, | |
| "loss": 0.3536, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3826955074875208, | |
| "grad_norm": 0.09427245200488364, | |
| "learning_rate": 0.00019953027957931658, | |
| "loss": 0.3621, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3860232945091514, | |
| "grad_norm": 0.09902024445411592, | |
| "learning_rate": 0.00019949198287790213, | |
| "loss": 0.3775, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.389351081530782, | |
| "grad_norm": 0.09515913211749447, | |
| "learning_rate": 0.00019945218953682734, | |
| "loss": 0.3692, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.39267886855241263, | |
| "grad_norm": 0.08817607353084106, | |
| "learning_rate": 0.00019941090015469613, | |
| "loss": 0.3596, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.39600665557404324, | |
| "grad_norm": 0.09420692817312656, | |
| "learning_rate": 0.00019936811535261714, | |
| "loss": 0.3633, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.39933444259567386, | |
| "grad_norm": 0.09046780862175126, | |
| "learning_rate": 0.00019932383577419432, | |
| "loss": 0.3567, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.40266222961730447, | |
| "grad_norm": 0.09175403219419788, | |
| "learning_rate": 0.00019927806208551717, | |
| "loss": 0.3584, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4059900166389351, | |
| "grad_norm": 0.09254222968074244, | |
| "learning_rate": 0.00019923079497515098, | |
| "loss": 0.3577, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.40931780366056575, | |
| "grad_norm": 0.09304434947638669, | |
| "learning_rate": 0.00019918203515412617, | |
| "loss": 0.3571, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.41264559068219636, | |
| "grad_norm": 0.09166662457636465, | |
| "learning_rate": 0.00019913178335592783, | |
| "loss": 0.351, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.415973377703827, | |
| "grad_norm": 0.09770977566631674, | |
| "learning_rate": 0.00019908004033648453, | |
| "loss": 0.3603, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4193011647254576, | |
| "grad_norm": 0.08989176229234969, | |
| "learning_rate": 0.00019902680687415705, | |
| "loss": 0.3635, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4226289517470882, | |
| "grad_norm": 0.10575374846231729, | |
| "learning_rate": 0.00019897208376972657, | |
| "loss": 0.3567, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.4259567387687188, | |
| "grad_norm": 0.09605030362372845, | |
| "learning_rate": 0.00019891587184638272, | |
| "loss": 0.3533, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4292845257903494, | |
| "grad_norm": 0.09421585492085938, | |
| "learning_rate": 0.00019885817194971117, | |
| "loss": 0.3565, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.43261231281198004, | |
| "grad_norm": 0.10039737771889443, | |
| "learning_rate": 0.00019879898494768093, | |
| "loss": 0.3625, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.43594009983361065, | |
| "grad_norm": 0.09522986889402585, | |
| "learning_rate": 0.00019873831173063114, | |
| "loss": 0.352, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.43926788685524126, | |
| "grad_norm": 0.10257926925753842, | |
| "learning_rate": 0.00019867615321125795, | |
| "loss": 0.3567, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4425956738768719, | |
| "grad_norm": 0.09288725590754815, | |
| "learning_rate": 0.0001986125103246005, | |
| "loss": 0.3474, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4459234608985025, | |
| "grad_norm": 0.10248152229336914, | |
| "learning_rate": 0.00019854738402802716, | |
| "loss": 0.3658, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4492512479201331, | |
| "grad_norm": 0.09962363019392478, | |
| "learning_rate": 0.00019848077530122083, | |
| "loss": 0.363, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4525790349417637, | |
| "grad_norm": 0.09514130939308267, | |
| "learning_rate": 0.00019841268514616433, | |
| "loss": 0.342, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4559068219633943, | |
| "grad_norm": 0.09623254569889135, | |
| "learning_rate": 0.00019834311458712545, | |
| "loss": 0.3436, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.45923460898502494, | |
| "grad_norm": 0.0963180375047454, | |
| "learning_rate": 0.00019827206467064133, | |
| "loss": 0.3605, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.46256239600665555, | |
| "grad_norm": 0.08780611129374004, | |
| "learning_rate": 0.00019819953646550286, | |
| "loss": 0.3613, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.46589018302828616, | |
| "grad_norm": 0.0932661436796585, | |
| "learning_rate": 0.00019812553106273847, | |
| "loss": 0.3506, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.46921797004991683, | |
| "grad_norm": 0.09994018909436361, | |
| "learning_rate": 0.00019805004957559793, | |
| "loss": 0.343, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.47254575707154745, | |
| "grad_norm": 0.09429430213876014, | |
| "learning_rate": 0.00019797309313953541, | |
| "loss": 0.3636, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.47587354409317806, | |
| "grad_norm": 0.09472120765392429, | |
| "learning_rate": 0.00019789466291219245, | |
| "loss": 0.358, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.47920133111480867, | |
| "grad_norm": 0.1005475330642602, | |
| "learning_rate": 0.00019781476007338058, | |
| "loss": 0.3626, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4825291181364393, | |
| "grad_norm": 0.0886723167191428, | |
| "learning_rate": 0.0001977333858250636, | |
| "loss": 0.3578, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4858569051580699, | |
| "grad_norm": 0.09134637666131144, | |
| "learning_rate": 0.0001976505413913393, | |
| "loss": 0.364, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4891846921797005, | |
| "grad_norm": 0.09424089868848846, | |
| "learning_rate": 0.00019756622801842143, | |
| "loss": 0.3555, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4925124792013311, | |
| "grad_norm": 0.09000532130097723, | |
| "learning_rate": 0.00019748044697462058, | |
| "loss": 0.3596, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.49584026622296173, | |
| "grad_norm": 0.09039755244703765, | |
| "learning_rate": 0.00019739319955032524, | |
| "loss": 0.3495, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.49916805324459235, | |
| "grad_norm": 0.10003839381412545, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.3636, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.502495840266223, | |
| "grad_norm": 0.09384514948321386, | |
| "learning_rate": 0.00019721431083207788, | |
| "loss": 0.3524, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5058236272878536, | |
| "grad_norm": 0.08795802790159053, | |
| "learning_rate": 0.00019712267222911606, | |
| "loss": 0.3599, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5091514143094842, | |
| "grad_norm": 0.08903283201308711, | |
| "learning_rate": 0.00019702957262759965, | |
| "loss": 0.3664, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5124792013311148, | |
| "grad_norm": 0.09295322271091727, | |
| "learning_rate": 0.00019693501342800892, | |
| "loss": 0.3461, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5158069883527454, | |
| "grad_norm": 0.09243114345245493, | |
| "learning_rate": 0.0001968389960527806, | |
| "loss": 0.3505, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.519134775374376, | |
| "grad_norm": 0.08538602702271118, | |
| "learning_rate": 0.00019674152194628638, | |
| "loss": 0.3493, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5224625623960066, | |
| "grad_norm": 0.09432000067886086, | |
| "learning_rate": 0.00019664259257481148, | |
| "loss": 0.3498, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5257903494176372, | |
| "grad_norm": 0.09670832759783635, | |
| "learning_rate": 0.0001965422094265322, | |
| "loss": 0.3656, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5291181364392679, | |
| "grad_norm": 0.0859953348514784, | |
| "learning_rate": 0.0001964403740114939, | |
| "loss": 0.349, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5324459234608985, | |
| "grad_norm": 0.08854229182039577, | |
| "learning_rate": 0.00019633708786158806, | |
| "loss": 0.3497, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5357737104825291, | |
| "grad_norm": 0.09412167167222708, | |
| "learning_rate": 0.00019623235253052922, | |
| "loss": 0.344, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5391014975041597, | |
| "grad_norm": 0.09390785580722581, | |
| "learning_rate": 0.0001961261695938319, | |
| "loss": 0.3458, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5424292845257903, | |
| "grad_norm": 0.09101659356452141, | |
| "learning_rate": 0.00019601854064878643, | |
| "loss": 0.3512, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5457570715474209, | |
| "grad_norm": 0.09273030612938668, | |
| "learning_rate": 0.0001959094673144354, | |
| "loss": 0.356, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5490848585690515, | |
| "grad_norm": 0.08915240514603591, | |
| "learning_rate": 0.0001957989512315489, | |
| "loss": 0.3511, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5524126455906821, | |
| "grad_norm": 0.09891272703781369, | |
| "learning_rate": 0.00019568699406260015, | |
| "loss": 0.346, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5557404326123128, | |
| "grad_norm": 0.08628455185678562, | |
| "learning_rate": 0.00019557359749174032, | |
| "loss": 0.3408, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5590682196339434, | |
| "grad_norm": 0.08910721810385827, | |
| "learning_rate": 0.0001954587632247732, | |
| "loss": 0.3538, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.562396006655574, | |
| "grad_norm": 0.0893630416494089, | |
| "learning_rate": 0.00019534249298912965, | |
| "loss": 0.3566, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5657237936772047, | |
| "grad_norm": 0.09157668416948159, | |
| "learning_rate": 0.00019522478853384155, | |
| "loss": 0.357, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5690515806988353, | |
| "grad_norm": 0.09738474969907182, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.3437, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5723793677204659, | |
| "grad_norm": 0.09053205783324851, | |
| "learning_rate": 0.00019498508406830576, | |
| "loss": 0.3435, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5757071547420965, | |
| "grad_norm": 0.08983029424441832, | |
| "learning_rate": 0.00019486308766388844, | |
| "loss": 0.3436, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5790349417637272, | |
| "grad_norm": 0.0917967998629968, | |
| "learning_rate": 0.00019473966425143292, | |
| "loss": 0.35, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5823627287853578, | |
| "grad_norm": 0.08877583492670515, | |
| "learning_rate": 0.00019461481568757506, | |
| "loss": 0.3451, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5856905158069884, | |
| "grad_norm": 0.09638380391131772, | |
| "learning_rate": 0.0001944885438503888, | |
| "loss": 0.3633, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.589018302828619, | |
| "grad_norm": 0.08787596372713234, | |
| "learning_rate": 0.00019436085063935835, | |
| "loss": 0.3414, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5923460898502496, | |
| "grad_norm": 0.09256039504021737, | |
| "learning_rate": 0.00019423173797534924, | |
| "loss": 0.3442, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5956738768718802, | |
| "grad_norm": 0.08721148776842906, | |
| "learning_rate": 0.00019410120780057958, | |
| "loss": 0.3418, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5990016638935108, | |
| "grad_norm": 0.09853922562566517, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.3582, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6023294509151415, | |
| "grad_norm": 0.08686249073031665, | |
| "learning_rate": 0.0001938359027942184, | |
| "loss": 0.3482, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6056572379367721, | |
| "grad_norm": 0.08379623104299327, | |
| "learning_rate": 0.0001937011319535615, | |
| "loss": 0.3493, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6089850249584027, | |
| "grad_norm": 0.08768506126689131, | |
| "learning_rate": 0.00019356495158395315, | |
| "loss": 0.3563, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6123128119800333, | |
| "grad_norm": 0.09120346903050645, | |
| "learning_rate": 0.00019342736373392975, | |
| "loss": 0.3597, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6156405990016639, | |
| "grad_norm": 0.08607194056274328, | |
| "learning_rate": 0.0001932883704732001, | |
| "loss": 0.3477, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6189683860232945, | |
| "grad_norm": 0.08829146932004177, | |
| "learning_rate": 0.00019314797389261424, | |
| "loss": 0.3518, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6222961730449251, | |
| "grad_norm": 0.08495297129753976, | |
| "learning_rate": 0.00019300617610413232, | |
| "loss": 0.3461, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6256239600665557, | |
| "grad_norm": 0.09463401109794949, | |
| "learning_rate": 0.00019286297924079243, | |
| "loss": 0.3596, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6289517470881864, | |
| "grad_norm": 0.08722418235312777, | |
| "learning_rate": 0.00019271838545667876, | |
| "loss": 0.3512, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.632279534109817, | |
| "grad_norm": 0.08410090811362046, | |
| "learning_rate": 0.00019257239692688907, | |
| "loss": 0.3541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6356073211314476, | |
| "grad_norm": 0.08485560221460367, | |
| "learning_rate": 0.00019242501584750202, | |
| "loss": 0.34, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6389351081530782, | |
| "grad_norm": 0.08800861870642733, | |
| "learning_rate": 0.00019227624443554425, | |
| "loss": 0.3578, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6422628951747088, | |
| "grad_norm": 0.08661797495574272, | |
| "learning_rate": 0.00019212608492895678, | |
| "loss": 0.3463, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6455906821963394, | |
| "grad_norm": 0.08570690192160581, | |
| "learning_rate": 0.00019197453958656158, | |
| "loss": 0.335, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.64891846921797, | |
| "grad_norm": 0.08787097065070806, | |
| "learning_rate": 0.00019182161068802741, | |
| "loss": 0.337, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6522462562396006, | |
| "grad_norm": 0.09202668113195005, | |
| "learning_rate": 0.00019166730053383568, | |
| "loss": 0.3412, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6555740432612313, | |
| "grad_norm": 0.08753285210853544, | |
| "learning_rate": 0.00019151161144524578, | |
| "loss": 0.3481, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6589018302828619, | |
| "grad_norm": 0.08802786231426123, | |
| "learning_rate": 0.0001913545457642601, | |
| "loss": 0.3451, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6622296173044925, | |
| "grad_norm": 0.09271185045887928, | |
| "learning_rate": 0.0001911961058535889, | |
| "loss": 0.3565, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6655574043261231, | |
| "grad_norm": 0.08201910669276406, | |
| "learning_rate": 0.0001910362940966147, | |
| "loss": 0.3424, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6688851913477537, | |
| "grad_norm": 0.0839694946959059, | |
| "learning_rate": 0.00019087511289735644, | |
| "loss": 0.3485, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6722129783693843, | |
| "grad_norm": 0.08730827459461137, | |
| "learning_rate": 0.00019071256468043338, | |
| "loss": 0.3347, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6755407653910149, | |
| "grad_norm": 0.08542343951540743, | |
| "learning_rate": 0.0001905486518910286, | |
| "loss": 0.3487, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6788685524126455, | |
| "grad_norm": 0.08845552825557125, | |
| "learning_rate": 0.00019038337699485208, | |
| "loss": 0.345, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6821963394342762, | |
| "grad_norm": 0.09105264577349556, | |
| "learning_rate": 0.0001902167424781038, | |
| "loss": 0.3411, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6855241264559068, | |
| "grad_norm": 0.09155791059326887, | |
| "learning_rate": 0.00019004875084743624, | |
| "loss": 0.354, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6888519134775375, | |
| "grad_norm": 0.08728918941188091, | |
| "learning_rate": 0.0001898794046299167, | |
| "loss": 0.3433, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6921797004991681, | |
| "grad_norm": 0.09028957874042293, | |
| "learning_rate": 0.00018970870637298934, | |
| "loss": 0.3395, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6955074875207987, | |
| "grad_norm": 0.08651865598937959, | |
| "learning_rate": 0.0001895366586444367, | |
| "loss": 0.3431, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6988352745424293, | |
| "grad_norm": 0.08319585954133932, | |
| "learning_rate": 0.00018936326403234125, | |
| "loss": 0.3471, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7021630615640599, | |
| "grad_norm": 0.0866379194314306, | |
| "learning_rate": 0.0001891885251450463, | |
| "loss": 0.3473, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7054908485856906, | |
| "grad_norm": 0.0857401490774991, | |
| "learning_rate": 0.00018901244461111695, | |
| "loss": 0.3607, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7088186356073212, | |
| "grad_norm": 0.08672595762343274, | |
| "learning_rate": 0.00018883502507930042, | |
| "loss": 0.3582, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7121464226289518, | |
| "grad_norm": 0.09049623112720732, | |
| "learning_rate": 0.00018865626921848615, | |
| "loss": 0.3432, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7154742096505824, | |
| "grad_norm": 0.0811001836660832, | |
| "learning_rate": 0.00018847617971766577, | |
| "loss": 0.3395, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.718801996672213, | |
| "grad_norm": 0.08798571287254327, | |
| "learning_rate": 0.00018829475928589271, | |
| "loss": 0.3381, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7221297836938436, | |
| "grad_norm": 0.08597426094232821, | |
| "learning_rate": 0.0001881120106522412, | |
| "loss": 0.3583, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7254575707154742, | |
| "grad_norm": 0.0859944971956249, | |
| "learning_rate": 0.00018792793656576543, | |
| "loss": 0.3439, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7287853577371048, | |
| "grad_norm": 0.08255890245178991, | |
| "learning_rate": 0.0001877425397954582, | |
| "loss": 0.3402, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7321131447587355, | |
| "grad_norm": 0.08631460040189731, | |
| "learning_rate": 0.0001875558231302091, | |
| "loss": 0.3512, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7354409317803661, | |
| "grad_norm": 0.08365169197048855, | |
| "learning_rate": 0.00018736778937876268, | |
| "loss": 0.3499, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7387687188019967, | |
| "grad_norm": 0.08372920255844533, | |
| "learning_rate": 0.00018717844136967624, | |
| "loss": 0.3438, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7420965058236273, | |
| "grad_norm": 0.08429349132245323, | |
| "learning_rate": 0.00018698778195127714, | |
| "loss": 0.3384, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7454242928452579, | |
| "grad_norm": 0.08745378307524065, | |
| "learning_rate": 0.00018679581399162008, | |
| "loss": 0.3484, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7487520798668885, | |
| "grad_norm": 0.08393648396724915, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.343, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7520798668885191, | |
| "grad_norm": 0.08455724098490583, | |
| "learning_rate": 0.00018640796401912807, | |
| "loss": 0.3519, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7554076539101497, | |
| "grad_norm": 0.08117288871452354, | |
| "learning_rate": 0.00018621208784064911, | |
| "loss": 0.3401, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7587354409317804, | |
| "grad_norm": 0.08405288096099374, | |
| "learning_rate": 0.00018601491478953657, | |
| "loss": 0.3372, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.762063227953411, | |
| "grad_norm": 0.08728491482380737, | |
| "learning_rate": 0.00018581644783182837, | |
| "loss": 0.3435, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7653910149750416, | |
| "grad_norm": 0.09105802986890799, | |
| "learning_rate": 0.00018561668995302667, | |
| "loss": 0.3483, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7687188019966722, | |
| "grad_norm": 0.08535152503021423, | |
| "learning_rate": 0.00018541564415805258, | |
| "loss": 0.3427, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7720465890183028, | |
| "grad_norm": 0.08368575404729736, | |
| "learning_rate": 0.00018521331347120114, | |
| "loss": 0.3609, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7753743760399334, | |
| "grad_norm": 0.08673975611038715, | |
| "learning_rate": 0.00018500970093609582, | |
| "loss": 0.3403, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.778702163061564, | |
| "grad_norm": 0.08627703201383483, | |
| "learning_rate": 0.0001848048096156426, | |
| "loss": 0.3357, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7820299500831946, | |
| "grad_norm": 0.08098393230784287, | |
| "learning_rate": 0.0001845986425919841, | |
| "loss": 0.3385, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7853577371048253, | |
| "grad_norm": 0.08810830878002943, | |
| "learning_rate": 0.00018439120296645308, | |
| "loss": 0.3426, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7886855241264559, | |
| "grad_norm": 0.0821139460904952, | |
| "learning_rate": 0.00018418249385952575, | |
| "loss": 0.3351, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7920133111480865, | |
| "grad_norm": 0.08731887904994778, | |
| "learning_rate": 0.000183972518410775, | |
| "loss": 0.3435, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7953410981697171, | |
| "grad_norm": 0.08047031443853961, | |
| "learning_rate": 0.00018376127977882294, | |
| "loss": 0.3312, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7986688851913477, | |
| "grad_norm": 0.08536196885920218, | |
| "learning_rate": 0.00018354878114129367, | |
| "loss": 0.3304, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8019966722129783, | |
| "grad_norm": 0.08099344341361203, | |
| "learning_rate": 0.00018333502569476516, | |
| "loss": 0.3454, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8053244592346089, | |
| "grad_norm": 0.08412143689707131, | |
| "learning_rate": 0.00018312001665472146, | |
| "loss": 0.3454, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8086522462562395, | |
| "grad_norm": 0.08055277302188278, | |
| "learning_rate": 0.00018290375725550417, | |
| "loss": 0.3344, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8119800332778702, | |
| "grad_norm": 0.08395291356544204, | |
| "learning_rate": 0.00018268625075026375, | |
| "loss": 0.3433, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8153078202995009, | |
| "grad_norm": 0.08687950482596268, | |
| "learning_rate": 0.0001824675004109107, | |
| "loss": 0.347, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8186356073211315, | |
| "grad_norm": 0.08568064564655968, | |
| "learning_rate": 0.00018224750952806624, | |
| "loss": 0.3361, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8219633943427621, | |
| "grad_norm": 0.0956445989295128, | |
| "learning_rate": 0.0001820262814110129, | |
| "loss": 0.3482, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8252911813643927, | |
| "grad_norm": 0.0860274416046879, | |
| "learning_rate": 0.00018180381938764478, | |
| "loss": 0.3452, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8286189683860233, | |
| "grad_norm": 0.08290877709780717, | |
| "learning_rate": 0.00018158012680441723, | |
| "loss": 0.3363, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.831946755407654, | |
| "grad_norm": 0.08745923486232092, | |
| "learning_rate": 0.00018135520702629675, | |
| "loss": 0.3325, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8352745424292846, | |
| "grad_norm": 0.08385476398036157, | |
| "learning_rate": 0.00018112906343671043, | |
| "loss": 0.3359, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8386023294509152, | |
| "grad_norm": 0.08053397247578684, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.3364, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8419301164725458, | |
| "grad_norm": 0.08383004215878544, | |
| "learning_rate": 0.0001806731184488447, | |
| "loss": 0.3255, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8452579034941764, | |
| "grad_norm": 0.0899951044366273, | |
| "learning_rate": 0.00018044332390926223, | |
| "loss": 0.3408, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.848585690515807, | |
| "grad_norm": 0.08206108121568055, | |
| "learning_rate": 0.0001802123192755044, | |
| "loss": 0.3472, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8519134775374376, | |
| "grad_norm": 0.08166211085852958, | |
| "learning_rate": 0.0001799801080225316, | |
| "loss": 0.3298, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8552412645590682, | |
| "grad_norm": 0.08459871874030749, | |
| "learning_rate": 0.00017974669364345517, | |
| "loss": 0.3465, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8585690515806988, | |
| "grad_norm": 0.08238051605367638, | |
| "learning_rate": 0.0001795120796494848, | |
| "loss": 0.3377, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8618968386023295, | |
| "grad_norm": 0.07959641908714889, | |
| "learning_rate": 0.00017927626956987578, | |
| "loss": 0.3403, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.8652246256239601, | |
| "grad_norm": 0.0806991496674187, | |
| "learning_rate": 0.00017903926695187595, | |
| "loss": 0.3444, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8685524126455907, | |
| "grad_norm": 0.0835202380550692, | |
| "learning_rate": 0.00017880107536067218, | |
| "loss": 0.334, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8718801996672213, | |
| "grad_norm": 0.08135590628602873, | |
| "learning_rate": 0.000178561698379337, | |
| "loss": 0.34, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8752079866888519, | |
| "grad_norm": 0.08254130280304803, | |
| "learning_rate": 0.00017832113960877444, | |
| "loss": 0.3401, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8785357737104825, | |
| "grad_norm": 0.07991377775486803, | |
| "learning_rate": 0.00017807940266766593, | |
| "loss": 0.3418, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8818635607321131, | |
| "grad_norm": 0.08789110268124221, | |
| "learning_rate": 0.00017783649119241602, | |
| "loss": 0.3487, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8851913477537438, | |
| "grad_norm": 0.08067371282703484, | |
| "learning_rate": 0.00017759240883709744, | |
| "loss": 0.3332, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8885191347753744, | |
| "grad_norm": 0.08305827454234688, | |
| "learning_rate": 0.0001773471592733964, | |
| "loss": 0.3434, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.891846921797005, | |
| "grad_norm": 0.0852925707044432, | |
| "learning_rate": 0.00017710074619055707, | |
| "loss": 0.3301, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.8951747088186356, | |
| "grad_norm": 0.08443430495666933, | |
| "learning_rate": 0.00017685317329532634, | |
| "loss": 0.3372, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8985024958402662, | |
| "grad_norm": 0.0812135526680112, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.3279, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9018302828618968, | |
| "grad_norm": 0.08470732237793946, | |
| "learning_rate": 0.00017635456298185605, | |
| "loss": 0.3215, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9051580698835274, | |
| "grad_norm": 0.07921259508735956, | |
| "learning_rate": 0.00017610353306412007, | |
| "loss": 0.3367, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.908485856905158, | |
| "grad_norm": 0.0814537166218216, | |
| "learning_rate": 0.00017585135833488692, | |
| "loss": 0.3492, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9118136439267887, | |
| "grad_norm": 0.08247099780517096, | |
| "learning_rate": 0.0001755980425875748, | |
| "loss": 0.3463, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9151414309484193, | |
| "grad_norm": 0.08633031091330437, | |
| "learning_rate": 0.00017534358963276607, | |
| "loss": 0.3258, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9184692179700499, | |
| "grad_norm": 0.08339996504487726, | |
| "learning_rate": 0.00017508800329814995, | |
| "loss": 0.3267, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9217970049916805, | |
| "grad_norm": 0.0861609612339358, | |
| "learning_rate": 0.0001748312874284647, | |
| "loss": 0.3348, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9251247920133111, | |
| "grad_norm": 0.08283652390749235, | |
| "learning_rate": 0.00017457344588544017, | |
| "loss": 0.3354, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9284525790349417, | |
| "grad_norm": 0.08328042739537875, | |
| "learning_rate": 0.00017431448254773944, | |
| "loss": 0.333, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9317803660565723, | |
| "grad_norm": 0.08780338301969765, | |
| "learning_rate": 0.00017405440131090048, | |
| "loss": 0.3532, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9351081530782029, | |
| "grad_norm": 0.08032941587332595, | |
| "learning_rate": 0.00017379320608727764, | |
| "loss": 0.3377, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9384359400998337, | |
| "grad_norm": 0.08140027870576953, | |
| "learning_rate": 0.0001735309008059829, | |
| "loss": 0.3364, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9417637271214643, | |
| "grad_norm": 0.08176011847546834, | |
| "learning_rate": 0.00017326748941282638, | |
| "loss": 0.3379, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9450915141430949, | |
| "grad_norm": 0.08062704920266694, | |
| "learning_rate": 0.00017300297587025747, | |
| "loss": 0.3401, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.9484193011647255, | |
| "grad_norm": 0.08760457164978003, | |
| "learning_rate": 0.00017273736415730488, | |
| "loss": 0.3349, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9517470881863561, | |
| "grad_norm": 0.07898030499985947, | |
| "learning_rate": 0.00017247065826951695, | |
| "loss": 0.3224, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9550748752079867, | |
| "grad_norm": 0.08163897477551052, | |
| "learning_rate": 0.00017220286221890135, | |
| "loss": 0.338, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9584026622296173, | |
| "grad_norm": 0.08202548074983111, | |
| "learning_rate": 0.0001719339800338651, | |
| "loss": 0.3478, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.961730449251248, | |
| "grad_norm": 0.08431106199269928, | |
| "learning_rate": 0.00017166401575915361, | |
| "loss": 0.336, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9650582362728786, | |
| "grad_norm": 0.0826212145998074, | |
| "learning_rate": 0.00017139297345578994, | |
| "loss": 0.3455, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9683860232945092, | |
| "grad_norm": 0.07722815146928662, | |
| "learning_rate": 0.00017112085720101373, | |
| "loss": 0.337, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.9717138103161398, | |
| "grad_norm": 0.08102709293851723, | |
| "learning_rate": 0.00017084767108822, | |
| "loss": 0.3389, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9750415973377704, | |
| "grad_norm": 0.07899429684757021, | |
| "learning_rate": 0.00017057341922689735, | |
| "loss": 0.3293, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.978369384359401, | |
| "grad_norm": 0.07719566747326663, | |
| "learning_rate": 0.0001702981057425662, | |
| "loss": 0.329, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.9816971713810316, | |
| "grad_norm": 0.08014647087178345, | |
| "learning_rate": 0.00017002173477671686, | |
| "loss": 0.3291, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9850249584026622, | |
| "grad_norm": 0.08298397794164776, | |
| "learning_rate": 0.00016974431048674715, | |
| "loss": 0.3442, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9883527454242929, | |
| "grad_norm": 0.08191027286565068, | |
| "learning_rate": 0.00016946583704589973, | |
| "loss": 0.3345, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9916805324459235, | |
| "grad_norm": 0.07941063301706962, | |
| "learning_rate": 0.00016918631864319957, | |
| "loss": 0.326, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.9950083194675541, | |
| "grad_norm": 0.07908114868236453, | |
| "learning_rate": 0.0001689057594833908, | |
| "loss": 0.335, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.9983361064891847, | |
| "grad_norm": 0.0831202355986244, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.3374, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9983361064891847, | |
| "eval_loss": 0.33711040019989014, | |
| "eval_runtime": 77.1419, | |
| "eval_samples_per_second": 26.25, | |
| "eval_steps_per_second": 0.83, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0016638935108153, | |
| "grad_norm": 0.08172897477894388, | |
| "learning_rate": 0.00016834153578963973, | |
| "loss": 0.3357, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.004991680532446, | |
| "grad_norm": 0.07525756510081694, | |
| "learning_rate": 0.00016805787974321105, | |
| "loss": 0.3266, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.0083194675540765, | |
| "grad_norm": 0.07804186287369605, | |
| "learning_rate": 0.00016777319991457325, | |
| "loss": 0.3266, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.0116472545757071, | |
| "grad_norm": 0.07609477567298127, | |
| "learning_rate": 0.00016748750058611278, | |
| "loss": 0.3244, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.0149750415973378, | |
| "grad_norm": 0.0789710597292621, | |
| "learning_rate": 0.00016720078605555224, | |
| "loss": 0.3257, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0183028286189684, | |
| "grad_norm": 0.07694532524092682, | |
| "learning_rate": 0.00016691306063588583, | |
| "loss": 0.312, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.021630615640599, | |
| "grad_norm": 0.08107958162859039, | |
| "learning_rate": 0.00016662432865531426, | |
| "loss": 0.3075, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.0249584026622296, | |
| "grad_norm": 0.08247253749043124, | |
| "learning_rate": 0.00016633459445717974, | |
| "loss": 0.3089, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.0282861896838602, | |
| "grad_norm": 0.08177201188880731, | |
| "learning_rate": 0.00016604386239990078, | |
| "loss": 0.3161, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.0316139767054908, | |
| "grad_norm": 0.08325643684555338, | |
| "learning_rate": 0.0001657521368569064, | |
| "loss": 0.3217, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0349417637271214, | |
| "grad_norm": 0.09206530401949085, | |
| "learning_rate": 0.0001654594222165704, | |
| "loss": 0.3297, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.038269550748752, | |
| "grad_norm": 0.08319595659110299, | |
| "learning_rate": 0.00016516572288214552, | |
| "loss": 0.3143, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.0415973377703827, | |
| "grad_norm": 0.07899287579549975, | |
| "learning_rate": 0.00016487104327169702, | |
| "loss": 0.3164, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.0449251247920133, | |
| "grad_norm": 0.08528320200651829, | |
| "learning_rate": 0.00016457538781803623, | |
| "loss": 0.3234, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.0482529118136439, | |
| "grad_norm": 0.07968728934053097, | |
| "learning_rate": 0.00016427876096865394, | |
| "loss": 0.3295, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0515806988352745, | |
| "grad_norm": 0.08324022167079065, | |
| "learning_rate": 0.00016398116718565348, | |
| "loss": 0.3283, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.054908485856905, | |
| "grad_norm": 0.08221519188360775, | |
| "learning_rate": 0.0001636826109456836, | |
| "loss": 0.3203, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0582362728785357, | |
| "grad_norm": 0.08277793526868028, | |
| "learning_rate": 0.00016338309673987101, | |
| "loss": 0.3294, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.0615640599001663, | |
| "grad_norm": 0.08456025835321836, | |
| "learning_rate": 0.00016308262907375313, | |
| "loss": 0.3204, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.064891846921797, | |
| "grad_norm": 0.08815773412894332, | |
| "learning_rate": 0.00016278121246720987, | |
| "loss": 0.3261, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0682196339434276, | |
| "grad_norm": 0.08496640111697164, | |
| "learning_rate": 0.000162478851454396, | |
| "loss": 0.3224, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0715474209650582, | |
| "grad_norm": 0.0931595428179426, | |
| "learning_rate": 0.0001621755505836729, | |
| "loss": 0.3203, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.0748752079866888, | |
| "grad_norm": 0.08294701676864061, | |
| "learning_rate": 0.0001618713144175399, | |
| "loss": 0.3162, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.0782029950083194, | |
| "grad_norm": 0.08281049983079036, | |
| "learning_rate": 0.0001615661475325658, | |
| "loss": 0.3113, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.08153078202995, | |
| "grad_norm": 0.08294879255223839, | |
| "learning_rate": 0.0001612600545193203, | |
| "loss": 0.3227, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0848585690515806, | |
| "grad_norm": 0.09212043203731173, | |
| "learning_rate": 0.00016095303998230433, | |
| "loss": 0.3336, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.0881863560732112, | |
| "grad_norm": 0.08007684690006908, | |
| "learning_rate": 0.00016064510853988138, | |
| "loss": 0.3123, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.0915141430948418, | |
| "grad_norm": 0.08286840373953612, | |
| "learning_rate": 0.00016033626482420758, | |
| "loss": 0.3108, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.0948419301164725, | |
| "grad_norm": 0.0820893679970151, | |
| "learning_rate": 0.00016002651348116247, | |
| "loss": 0.3195, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.098169717138103, | |
| "grad_norm": 0.08632984297969504, | |
| "learning_rate": 0.00015971585917027862, | |
| "loss": 0.3265, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1014975041597337, | |
| "grad_norm": 0.08495381612420334, | |
| "learning_rate": 0.00015940430656467193, | |
| "loss": 0.3275, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.1048252911813643, | |
| "grad_norm": 0.08358801834908114, | |
| "learning_rate": 0.00015909186035097111, | |
| "loss": 0.3209, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.108153078202995, | |
| "grad_norm": 0.08420884068917062, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.3282, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.1114808652246255, | |
| "grad_norm": 0.08515518698668575, | |
| "learning_rate": 0.0001584643059129433, | |
| "loss": 0.3136, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.1148086522462561, | |
| "grad_norm": 0.08268658015951906, | |
| "learning_rate": 0.00015814920712880267, | |
| "loss": 0.3151, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1181364392678868, | |
| "grad_norm": 0.08220653725508614, | |
| "learning_rate": 0.00015783323361679864, | |
| "loss": 0.3222, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.1214642262895174, | |
| "grad_norm": 0.08291266698007108, | |
| "learning_rate": 0.0001575163901300629, | |
| "loss": 0.3046, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.124792013311148, | |
| "grad_norm": 0.0874811875094012, | |
| "learning_rate": 0.00015719868143481384, | |
| "loss": 0.3264, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.1281198003327786, | |
| "grad_norm": 0.08667448012302992, | |
| "learning_rate": 0.00015688011231028518, | |
| "loss": 0.3204, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.1314475873544092, | |
| "grad_norm": 0.08257673794327997, | |
| "learning_rate": 0.00015656068754865387, | |
| "loss": 0.3183, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1347753743760398, | |
| "grad_norm": 0.0819695134151146, | |
| "learning_rate": 0.000156240411954968, | |
| "loss": 0.3284, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.1381031613976704, | |
| "grad_norm": 0.08222215067678788, | |
| "learning_rate": 0.0001559192903470747, | |
| "loss": 0.3222, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.1414309484193013, | |
| "grad_norm": 0.08396024282192119, | |
| "learning_rate": 0.00015559732755554735, | |
| "loss": 0.3325, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.1447587354409319, | |
| "grad_norm": 0.08175389304427882, | |
| "learning_rate": 0.00015527452842361327, | |
| "loss": 0.3208, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.1480865224625625, | |
| "grad_norm": 0.07908598841046571, | |
| "learning_rate": 0.0001549508978070806, | |
| "loss": 0.308, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.151414309484193, | |
| "grad_norm": 0.08370444886296384, | |
| "learning_rate": 0.0001546264405742654, | |
| "loss": 0.3227, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.1547420965058237, | |
| "grad_norm": 0.08553465907055378, | |
| "learning_rate": 0.00015430116160591834, | |
| "loss": 0.3196, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.1580698835274543, | |
| "grad_norm": 0.08027798959427773, | |
| "learning_rate": 0.0001539750657951513, | |
| "loss": 0.326, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.161397670549085, | |
| "grad_norm": 0.08339428493778815, | |
| "learning_rate": 0.0001536481580473638, | |
| "loss": 0.3248, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.1647254575707155, | |
| "grad_norm": 0.08241224264426707, | |
| "learning_rate": 0.00015332044328016914, | |
| "loss": 0.3182, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1680532445923462, | |
| "grad_norm": 0.08480183966625279, | |
| "learning_rate": 0.0001529919264233205, | |
| "loss": 0.3118, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.1713810316139768, | |
| "grad_norm": 0.08381465549809634, | |
| "learning_rate": 0.00015266261241863674, | |
| "loss": 0.3196, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.1747088186356074, | |
| "grad_norm": 0.07989646959202727, | |
| "learning_rate": 0.0001523325062199281, | |
| "loss": 0.3126, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.178036605657238, | |
| "grad_norm": 0.08067827718704482, | |
| "learning_rate": 0.00015200161279292155, | |
| "loss": 0.3204, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.1813643926788686, | |
| "grad_norm": 0.08006890429745794, | |
| "learning_rate": 0.00015166993711518631, | |
| "loss": 0.3155, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.1846921797004992, | |
| "grad_norm": 0.0810834421876794, | |
| "learning_rate": 0.00015133748417605876, | |
| "loss": 0.3268, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.1880199667221298, | |
| "grad_norm": 0.07965698918148639, | |
| "learning_rate": 0.00015100425897656753, | |
| "loss": 0.3128, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.1913477537437605, | |
| "grad_norm": 0.08084392890696793, | |
| "learning_rate": 0.00015067026652935823, | |
| "loss": 0.3334, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.194675540765391, | |
| "grad_norm": 0.08047404773386269, | |
| "learning_rate": 0.000150335511858618, | |
| "loss": 0.3278, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.1980033277870217, | |
| "grad_norm": 0.08115839240152477, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.3156, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2013311148086523, | |
| "grad_norm": 0.08418442960560545, | |
| "learning_rate": 0.0001496637360005476, | |
| "loss": 0.3227, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.204658901830283, | |
| "grad_norm": 0.07908059971720892, | |
| "learning_rate": 0.00014932672491861854, | |
| "loss": 0.33, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.2079866888519135, | |
| "grad_norm": 0.08108209444129062, | |
| "learning_rate": 0.0001489889718238087, | |
| "loss": 0.3154, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.2113144758735441, | |
| "grad_norm": 0.08127947954186157, | |
| "learning_rate": 0.000148650481796876, | |
| "loss": 0.3271, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.2146422628951747, | |
| "grad_norm": 0.08432099836336952, | |
| "learning_rate": 0.00014831125992966385, | |
| "loss": 0.3253, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.2179700499168054, | |
| "grad_norm": 0.08234909676296243, | |
| "learning_rate": 0.00014797131132502465, | |
| "loss": 0.3165, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.221297836938436, | |
| "grad_norm": 0.08182985880054604, | |
| "learning_rate": 0.0001476306410967429, | |
| "loss": 0.3213, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.2246256239600666, | |
| "grad_norm": 0.08190124708255395, | |
| "learning_rate": 0.00014728925436945838, | |
| "loss": 0.3275, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.2279534109816972, | |
| "grad_norm": 0.08336634463763762, | |
| "learning_rate": 0.00014694715627858908, | |
| "loss": 0.3226, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.2312811980033278, | |
| "grad_norm": 0.0873959573836208, | |
| "learning_rate": 0.0001466043519702539, | |
| "loss": 0.3178, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2346089850249584, | |
| "grad_norm": 0.08045422800249838, | |
| "learning_rate": 0.00014626084660119515, | |
| "loss": 0.3109, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.237936772046589, | |
| "grad_norm": 0.0819888980325759, | |
| "learning_rate": 0.00014591664533870118, | |
| "loss": 0.3226, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.2412645590682196, | |
| "grad_norm": 0.0845779702064252, | |
| "learning_rate": 0.00014557175336052844, | |
| "loss": 0.325, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.2445923460898503, | |
| "grad_norm": 0.08779550722212086, | |
| "learning_rate": 0.00014522617585482377, | |
| "loss": 0.3175, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.2479201331114809, | |
| "grad_norm": 0.08196385855641611, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.3217, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2512479201331115, | |
| "grad_norm": 0.08138061511354558, | |
| "learning_rate": 0.00014453298506488896, | |
| "loss": 0.3199, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.254575707154742, | |
| "grad_norm": 0.08376705254870072, | |
| "learning_rate": 0.0001441853822082008, | |
| "loss": 0.3262, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.2579034941763727, | |
| "grad_norm": 0.08337076801897848, | |
| "learning_rate": 0.00014383711467890774, | |
| "loss": 0.31, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.2612312811980033, | |
| "grad_norm": 0.08294854116735015, | |
| "learning_rate": 0.0001434881877159345, | |
| "loss": 0.3267, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.264559068219634, | |
| "grad_norm": 0.07949371359379681, | |
| "learning_rate": 0.00014313860656812536, | |
| "loss": 0.3132, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2678868552412645, | |
| "grad_norm": 0.08245690284894774, | |
| "learning_rate": 0.00014278837649416544, | |
| "loss": 0.3152, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.2712146422628952, | |
| "grad_norm": 0.07758599936792739, | |
| "learning_rate": 0.00014243750276250153, | |
| "loss": 0.3084, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.2745424292845258, | |
| "grad_norm": 0.08262502240727569, | |
| "learning_rate": 0.00014208599065126292, | |
| "loss": 0.3337, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.2778702163061564, | |
| "grad_norm": 0.08262955750406191, | |
| "learning_rate": 0.0001417338454481818, | |
| "loss": 0.3265, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.281198003327787, | |
| "grad_norm": 0.08180077695711627, | |
| "learning_rate": 0.00014138107245051392, | |
| "loss": 0.3201, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.2845257903494176, | |
| "grad_norm": 0.07888524705775452, | |
| "learning_rate": 0.00014102767696495884, | |
| "loss": 0.3128, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.2878535773710482, | |
| "grad_norm": 0.08307087945812033, | |
| "learning_rate": 0.00014067366430758004, | |
| "loss": 0.3138, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.2911813643926788, | |
| "grad_norm": 0.08054315530573528, | |
| "learning_rate": 0.00014031903980372504, | |
| "loss": 0.3139, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.2945091514143094, | |
| "grad_norm": 0.07868497352148025, | |
| "learning_rate": 0.00013996380878794523, | |
| "loss": 0.3106, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.29783693843594, | |
| "grad_norm": 0.08247651282903071, | |
| "learning_rate": 0.0001396079766039157, | |
| "loss": 0.3162, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3011647254575707, | |
| "grad_norm": 0.08216756230606265, | |
| "learning_rate": 0.00013925154860435472, | |
| "loss": 0.3163, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.3044925124792013, | |
| "grad_norm": 0.08185506485000643, | |
| "learning_rate": 0.00013889453015094338, | |
| "loss": 0.3197, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.307820299500832, | |
| "grad_norm": 0.07856586629145418, | |
| "learning_rate": 0.00013853692661424484, | |
| "loss": 0.3198, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.3111480865224625, | |
| "grad_norm": 0.08060967720700726, | |
| "learning_rate": 0.0001381787433736235, | |
| "loss": 0.3225, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.3144758735440931, | |
| "grad_norm": 0.08107785255016375, | |
| "learning_rate": 0.00013781998581716427, | |
| "loss": 0.3221, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3178036605657237, | |
| "grad_norm": 0.08129218944467384, | |
| "learning_rate": 0.00013746065934159123, | |
| "loss": 0.323, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.3211314475873543, | |
| "grad_norm": 0.08024616882621527, | |
| "learning_rate": 0.0001371007693521867, | |
| "loss": 0.3246, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.324459234608985, | |
| "grad_norm": 0.0802781251175933, | |
| "learning_rate": 0.00013674032126270982, | |
| "loss": 0.3265, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.3277870216306156, | |
| "grad_norm": 0.07981140634199578, | |
| "learning_rate": 0.00013637932049531516, | |
| "loss": 0.3245, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.3311148086522462, | |
| "grad_norm": 0.08648826065841989, | |
| "learning_rate": 0.00013601777248047105, | |
| "loss": 0.3471, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3344425956738768, | |
| "grad_norm": 0.08317523942952117, | |
| "learning_rate": 0.000135655682656878, | |
| "loss": 0.3178, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.3377703826955074, | |
| "grad_norm": 0.07996211762258155, | |
| "learning_rate": 0.00013529305647138687, | |
| "loss": 0.2972, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.341098169717138, | |
| "grad_norm": 0.08903676659771695, | |
| "learning_rate": 0.00013492989937891693, | |
| "loss": 0.3184, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.3444259567387689, | |
| "grad_norm": 0.08501846581599046, | |
| "learning_rate": 0.00013456621684237367, | |
| "loss": 0.3239, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.3477537437603995, | |
| "grad_norm": 0.0772390839406517, | |
| "learning_rate": 0.00013420201433256689, | |
| "loss": 0.321, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.35108153078203, | |
| "grad_norm": 0.08051701684207763, | |
| "learning_rate": 0.00013383729732812814, | |
| "loss": 0.3141, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.3544093178036607, | |
| "grad_norm": 0.08298731698275244, | |
| "learning_rate": 0.00013347207131542848, | |
| "loss": 0.3182, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.3577371048252913, | |
| "grad_norm": 0.08527256051541643, | |
| "learning_rate": 0.0001331063417884958, | |
| "loss": 0.3176, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.361064891846922, | |
| "grad_norm": 0.08077817536970006, | |
| "learning_rate": 0.00013274011424893245, | |
| "loss": 0.3048, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.3643926788685525, | |
| "grad_norm": 0.07677673810294505, | |
| "learning_rate": 0.00013237339420583212, | |
| "loss": 0.3183, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3677204658901831, | |
| "grad_norm": 0.07972594016931787, | |
| "learning_rate": 0.00013200618717569714, | |
| "loss": 0.3115, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.3710482529118138, | |
| "grad_norm": 0.08092858982570945, | |
| "learning_rate": 0.00013163849868235564, | |
| "loss": 0.3202, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.3743760399334444, | |
| "grad_norm": 0.07944845432009648, | |
| "learning_rate": 0.0001312703342568782, | |
| "loss": 0.3248, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.377703826955075, | |
| "grad_norm": 0.0780652305156278, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.3092, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.3810316139767056, | |
| "grad_norm": 0.08414709507839971, | |
| "learning_rate": 0.00013053259976951133, | |
| "loss": 0.3271, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.3843594009983362, | |
| "grad_norm": 0.0847046966062323, | |
| "learning_rate": 0.00013016304080522656, | |
| "loss": 0.327, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.3876871880199668, | |
| "grad_norm": 0.08100848693932221, | |
| "learning_rate": 0.0001297930281038482, | |
| "loss": 0.31, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.3910149750415974, | |
| "grad_norm": 0.08241037415049582, | |
| "learning_rate": 0.00012942256723140952, | |
| "loss": 0.3133, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.394342762063228, | |
| "grad_norm": 0.0835290106718076, | |
| "learning_rate": 0.0001290516637606855, | |
| "loss": 0.3071, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.3976705490848587, | |
| "grad_norm": 0.08292584699938567, | |
| "learning_rate": 0.00012868032327110904, | |
| "loss": 0.3147, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4009983361064893, | |
| "grad_norm": 0.07828614710520194, | |
| "learning_rate": 0.00012830855134868705, | |
| "loss": 0.3166, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.4043261231281199, | |
| "grad_norm": 0.08174730218560904, | |
| "learning_rate": 0.00012793635358591645, | |
| "loss": 0.3123, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.4076539101497505, | |
| "grad_norm": 0.07767874535447017, | |
| "learning_rate": 0.0001275637355816999, | |
| "loss": 0.3191, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.410981697171381, | |
| "grad_norm": 0.08086800064329293, | |
| "learning_rate": 0.00012719070294126182, | |
| "loss": 0.3271, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.4143094841930117, | |
| "grad_norm": 0.07497662640393904, | |
| "learning_rate": 0.00012681726127606376, | |
| "loss": 0.3309, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4176372712146423, | |
| "grad_norm": 0.07962764783000117, | |
| "learning_rate": 0.00012644341620372023, | |
| "loss": 0.3191, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.420965058236273, | |
| "grad_norm": 0.07861027348332184, | |
| "learning_rate": 0.00012606917334791415, | |
| "loss": 0.3157, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.4242928452579036, | |
| "grad_norm": 0.07965374328815958, | |
| "learning_rate": 0.00012569453833831222, | |
| "loss": 0.3214, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.4276206322795342, | |
| "grad_norm": 0.08400097497468106, | |
| "learning_rate": 0.0001253195168104802, | |
| "loss": 0.3175, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.4309484193011648, | |
| "grad_norm": 0.08084295786403127, | |
| "learning_rate": 0.00012494411440579814, | |
| "loss": 0.3235, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4342762063227954, | |
| "grad_norm": 0.08122836356004892, | |
| "learning_rate": 0.00012456833677137563, | |
| "loss": 0.314, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.437603993344426, | |
| "grad_norm": 0.08280103608529468, | |
| "learning_rate": 0.00012419218955996676, | |
| "loss": 0.3171, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.4409317803660566, | |
| "grad_norm": 0.08205503810732262, | |
| "learning_rate": 0.0001238156784298851, | |
| "loss": 0.3254, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.4442595673876872, | |
| "grad_norm": 0.079419735374231, | |
| "learning_rate": 0.00012343880904491848, | |
| "loss": 0.3187, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.4475873544093179, | |
| "grad_norm": 0.07828839386614656, | |
| "learning_rate": 0.00012306158707424403, | |
| "loss": 0.3198, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4509151414309485, | |
| "grad_norm": 0.07907038788844414, | |
| "learning_rate": 0.0001226840181923427, | |
| "loss": 0.3197, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.454242928452579, | |
| "grad_norm": 0.08325826886590487, | |
| "learning_rate": 0.00012230610807891393, | |
| "loss": 0.314, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.4575707154742097, | |
| "grad_norm": 0.0806247097494845, | |
| "learning_rate": 0.00012192786241879033, | |
| "loss": 0.3187, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.4608985024958403, | |
| "grad_norm": 0.07724581172198981, | |
| "learning_rate": 0.000121549286901852, | |
| "loss": 0.3097, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.464226289517471, | |
| "grad_norm": 0.07899457825791725, | |
| "learning_rate": 0.0001211703872229411, | |
| "loss": 0.3139, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4675540765391015, | |
| "grad_norm": 0.08047079959840484, | |
| "learning_rate": 0.00012079116908177593, | |
| "loss": 0.315, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.4708818635607321, | |
| "grad_norm": 0.08420613982148384, | |
| "learning_rate": 0.00012041163818286559, | |
| "loss": 0.3207, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.4742096505823628, | |
| "grad_norm": 0.07678473115460477, | |
| "learning_rate": 0.00012003180023542375, | |
| "loss": 0.3077, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.4775374376039934, | |
| "grad_norm": 0.07973964616933353, | |
| "learning_rate": 0.00011965166095328301, | |
| "loss": 0.3194, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.480865224625624, | |
| "grad_norm": 0.0790069302947316, | |
| "learning_rate": 0.00011927122605480898, | |
| "loss": 0.3189, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.4841930116472546, | |
| "grad_norm": 0.08153773044879684, | |
| "learning_rate": 0.00011889050126281405, | |
| "loss": 0.3218, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.4875207986688852, | |
| "grad_norm": 0.08501103953134748, | |
| "learning_rate": 0.00011850949230447145, | |
| "loss": 0.3137, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.4908485856905158, | |
| "grad_norm": 0.07309681155028981, | |
| "learning_rate": 0.00011812820491122918, | |
| "loss": 0.2995, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.4941763727121464, | |
| "grad_norm": 0.08015131369228977, | |
| "learning_rate": 0.00011774664481872353, | |
| "loss": 0.3253, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.497504159733777, | |
| "grad_norm": 0.07738915299948486, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.323, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.5008319467554077, | |
| "grad_norm": 0.08185934914432139, | |
| "learning_rate": 0.00011698272949889207, | |
| "loss": 0.3082, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.5041597337770383, | |
| "grad_norm": 0.08184414118465821, | |
| "learning_rate": 0.00011660038576300443, | |
| "loss": 0.3269, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.5074875207986689, | |
| "grad_norm": 0.07604076302522539, | |
| "learning_rate": 0.00011621779231055676, | |
| "loss": 0.3108, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.5108153078202995, | |
| "grad_norm": 0.0785967501755232, | |
| "learning_rate": 0.00011583495489683229, | |
| "loss": 0.333, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.51414309484193, | |
| "grad_norm": 0.0808202397707487, | |
| "learning_rate": 0.00011545187928078406, | |
| "loss": 0.3192, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.5174708818635607, | |
| "grad_norm": 0.0798366081778451, | |
| "learning_rate": 0.00011506857122494831, | |
| "loss": 0.3172, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.5207986688851913, | |
| "grad_norm": 0.07521275253640944, | |
| "learning_rate": 0.00011468503649535789, | |
| "loss": 0.2968, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.524126455906822, | |
| "grad_norm": 0.07996600324418522, | |
| "learning_rate": 0.00011430128086145542, | |
| "loss": 0.3225, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.5274542429284526, | |
| "grad_norm": 0.07963407083209063, | |
| "learning_rate": 0.00011391731009600654, | |
| "loss": 0.3164, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.5307820299500832, | |
| "grad_norm": 0.08265056954458613, | |
| "learning_rate": 0.00011353312997501313, | |
| "loss": 0.3134, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5341098169717138, | |
| "grad_norm": 0.07933690023360569, | |
| "learning_rate": 0.00011314874627762627, | |
| "loss": 0.3165, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.5374376039933444, | |
| "grad_norm": 0.08177560529225479, | |
| "learning_rate": 0.00011276416478605949, | |
| "loss": 0.3131, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.540765391014975, | |
| "grad_norm": 0.08189925840022073, | |
| "learning_rate": 0.00011237939128550166, | |
| "loss": 0.325, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.5440931780366056, | |
| "grad_norm": 0.08168307893652589, | |
| "learning_rate": 0.00011199443156402998, | |
| "loss": 0.3251, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.5474209650582362, | |
| "grad_norm": 0.08242480069785535, | |
| "learning_rate": 0.00011160929141252303, | |
| "loss": 0.3297, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5507487520798668, | |
| "grad_norm": 0.0809784878320606, | |
| "learning_rate": 0.0001112239766245735, | |
| "loss": 0.3208, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.5540765391014975, | |
| "grad_norm": 0.07862074249847724, | |
| "learning_rate": 0.00011083849299640108, | |
| "loss": 0.3164, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.557404326123128, | |
| "grad_norm": 0.07767243998434667, | |
| "learning_rate": 0.00011045284632676536, | |
| "loss": 0.309, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.5607321131447587, | |
| "grad_norm": 0.0806548614296814, | |
| "learning_rate": 0.00011006704241687847, | |
| "loss": 0.3112, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.5640599001663893, | |
| "grad_norm": 0.07993563511207265, | |
| "learning_rate": 0.00010968108707031792, | |
| "loss": 0.3097, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.56738768718802, | |
| "grad_norm": 0.07991539122406761, | |
| "learning_rate": 0.00010929498609293924, | |
| "loss": 0.3089, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.5707154742096505, | |
| "grad_norm": 0.08282772536663943, | |
| "learning_rate": 0.00010890874529278865, | |
| "loss": 0.3281, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.5740432612312811, | |
| "grad_norm": 0.08226398967982722, | |
| "learning_rate": 0.00010852237048001567, | |
| "loss": 0.3115, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.5773710482529117, | |
| "grad_norm": 0.08155207338763963, | |
| "learning_rate": 0.00010813586746678583, | |
| "loss": 0.3141, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.5806988352745424, | |
| "grad_norm": 0.07626735480523346, | |
| "learning_rate": 0.0001077492420671931, | |
| "loss": 0.3146, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.584026622296173, | |
| "grad_norm": 0.08184353491174834, | |
| "learning_rate": 0.00010736250009717247, | |
| "loss": 0.3199, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.5873544093178036, | |
| "grad_norm": 0.08138099188835729, | |
| "learning_rate": 0.00010697564737441252, | |
| "loss": 0.3198, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.5906821963394342, | |
| "grad_norm": 0.07714660939538655, | |
| "learning_rate": 0.00010658868971826785, | |
| "loss": 0.3091, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.5940099833610648, | |
| "grad_norm": 0.0782409788482085, | |
| "learning_rate": 0.00010620163294967154, | |
| "loss": 0.3195, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.5973377703826954, | |
| "grad_norm": 0.07633338509194124, | |
| "learning_rate": 0.00010581448289104758, | |
| "loss": 0.3059, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.600665557404326, | |
| "grad_norm": 0.08051801433773054, | |
| "learning_rate": 0.0001054272453662234, | |
| "loss": 0.3214, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.6039933444259566, | |
| "grad_norm": 0.07887049158194588, | |
| "learning_rate": 0.00010503992620034202, | |
| "loss": 0.3094, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.6073211314475873, | |
| "grad_norm": 0.07842159146503913, | |
| "learning_rate": 0.0001046525312197747, | |
| "loss": 0.3107, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.6106489184692179, | |
| "grad_norm": 0.07770313377756956, | |
| "learning_rate": 0.00010426506625203307, | |
| "loss": 0.3186, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.6139767054908485, | |
| "grad_norm": 0.08054293517894129, | |
| "learning_rate": 0.0001038775371256817, | |
| "loss": 0.3328, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.617304492512479, | |
| "grad_norm": 0.07935603869302343, | |
| "learning_rate": 0.00010348994967025012, | |
| "loss": 0.3278, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.6206322795341097, | |
| "grad_norm": 0.07889873664213935, | |
| "learning_rate": 0.00010310230971614538, | |
| "loss": 0.3118, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.6239600665557403, | |
| "grad_norm": 0.08065624118328649, | |
| "learning_rate": 0.0001027146230945643, | |
| "loss": 0.3125, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.627287853577371, | |
| "grad_norm": 0.07936759162012692, | |
| "learning_rate": 0.00010232689563740563, | |
| "loss": 0.3269, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.6306156405990015, | |
| "grad_norm": 0.0828976936543311, | |
| "learning_rate": 0.00010193913317718244, | |
| "loss": 0.3223, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6339434276206322, | |
| "grad_norm": 0.07857900363659823, | |
| "learning_rate": 0.00010155134154693433, | |
| "loss": 0.3246, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.6372712146422628, | |
| "grad_norm": 0.08085870450710857, | |
| "learning_rate": 0.00010116352658013973, | |
| "loss": 0.3126, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.6405990016638934, | |
| "grad_norm": 0.08075462984614992, | |
| "learning_rate": 0.00010077569411062803, | |
| "loss": 0.3106, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.643926788685524, | |
| "grad_norm": 0.0808327847718592, | |
| "learning_rate": 0.00010038784997249205, | |
| "loss": 0.3153, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.6472545757071546, | |
| "grad_norm": 0.0794147040861997, | |
| "learning_rate": 0.0001, | |
| "loss": 0.311, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.6505823627287852, | |
| "grad_norm": 0.07916725392881477, | |
| "learning_rate": 9.961215002750799e-05, | |
| "loss": 0.304, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.6539101497504158, | |
| "grad_norm": 0.07779521029572718, | |
| "learning_rate": 9.922430588937199e-05, | |
| "loss": 0.3186, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.6572379367720464, | |
| "grad_norm": 0.08015313045412313, | |
| "learning_rate": 9.883647341986032e-05, | |
| "loss": 0.3105, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.660565723793677, | |
| "grad_norm": 0.07871411311807877, | |
| "learning_rate": 9.844865845306568e-05, | |
| "loss": 0.3091, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.6638935108153077, | |
| "grad_norm": 0.07830337205751042, | |
| "learning_rate": 9.806086682281758e-05, | |
| "loss": 0.3111, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6672212978369383, | |
| "grad_norm": 0.07733539972336034, | |
| "learning_rate": 9.767310436259438e-05, | |
| "loss": 0.3089, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.670549084858569, | |
| "grad_norm": 0.07979774444110035, | |
| "learning_rate": 9.728537690543572e-05, | |
| "loss": 0.3146, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.6738768718801995, | |
| "grad_norm": 0.08067446292443037, | |
| "learning_rate": 9.689769028385463e-05, | |
| "loss": 0.319, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.6772046589018301, | |
| "grad_norm": 0.07784440875324662, | |
| "learning_rate": 9.651005032974994e-05, | |
| "loss": 0.3222, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.6805324459234607, | |
| "grad_norm": 0.07861786512866405, | |
| "learning_rate": 9.612246287431831e-05, | |
| "loss": 0.3161, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.6838602329450914, | |
| "grad_norm": 0.0804699829788408, | |
| "learning_rate": 9.573493374796693e-05, | |
| "loss": 0.3196, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.687188019966722, | |
| "grad_norm": 0.07906127678258533, | |
| "learning_rate": 9.534746878022534e-05, | |
| "loss": 0.3181, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.6905158069883528, | |
| "grad_norm": 0.07724902426755939, | |
| "learning_rate": 9.496007379965801e-05, | |
| "loss": 0.3087, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.6938435940099834, | |
| "grad_norm": 0.07632495487523883, | |
| "learning_rate": 9.457275463377664e-05, | |
| "loss": 0.3245, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.697171381031614, | |
| "grad_norm": 0.07883577314074905, | |
| "learning_rate": 9.418551710895243e-05, | |
| "loss": 0.3276, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.7004991680532446, | |
| "grad_norm": 0.08068776992303811, | |
| "learning_rate": 9.379836705032848e-05, | |
| "loss": 0.32, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.7038269550748752, | |
| "grad_norm": 0.07537204047501166, | |
| "learning_rate": 9.341131028173214e-05, | |
| "loss": 0.3044, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.7071547420965059, | |
| "grad_norm": 0.08275383774251882, | |
| "learning_rate": 9.302435262558747e-05, | |
| "loss": 0.3295, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.7104825291181365, | |
| "grad_norm": 0.07972947554756316, | |
| "learning_rate": 9.263749990282754e-05, | |
| "loss": 0.3241, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.713810316139767, | |
| "grad_norm": 0.07819408735705255, | |
| "learning_rate": 9.225075793280692e-05, | |
| "loss": 0.3198, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.7171381031613977, | |
| "grad_norm": 0.07757269688260389, | |
| "learning_rate": 9.186413253321418e-05, | |
| "loss": 0.3087, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.7204658901830283, | |
| "grad_norm": 0.0803685738978913, | |
| "learning_rate": 9.147762951998435e-05, | |
| "loss": 0.3158, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.723793677204659, | |
| "grad_norm": 0.07814934088541807, | |
| "learning_rate": 9.10912547072114e-05, | |
| "loss": 0.3143, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.7271214642262895, | |
| "grad_norm": 0.07801931480494577, | |
| "learning_rate": 9.070501390706079e-05, | |
| "loss": 0.3042, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.7304492512479202, | |
| "grad_norm": 0.0796930726792969, | |
| "learning_rate": 9.03189129296821e-05, | |
| "loss": 0.3253, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7337770382695508, | |
| "grad_norm": 0.07869331844450607, | |
| "learning_rate": 8.993295758312155e-05, | |
| "loss": 0.3042, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.7371048252911814, | |
| "grad_norm": 0.0783622077737167, | |
| "learning_rate": 8.954715367323468e-05, | |
| "loss": 0.3187, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.740432612312812, | |
| "grad_norm": 0.08096765714219455, | |
| "learning_rate": 8.916150700359896e-05, | |
| "loss": 0.326, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.7437603993344426, | |
| "grad_norm": 0.07896182979041651, | |
| "learning_rate": 8.877602337542655e-05, | |
| "loss": 0.3144, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.7470881863560732, | |
| "grad_norm": 0.07967420319709191, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.3219, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7504159733777038, | |
| "grad_norm": 0.077411618341554, | |
| "learning_rate": 8.800556843597002e-05, | |
| "loss": 0.3142, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.7537437603993344, | |
| "grad_norm": 0.07808504060370658, | |
| "learning_rate": 8.762060871449838e-05, | |
| "loss": 0.2985, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.757071547420965, | |
| "grad_norm": 0.07812149609981953, | |
| "learning_rate": 8.723583521394054e-05, | |
| "loss": 0.3098, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.7603993344425957, | |
| "grad_norm": 0.08061195839845843, | |
| "learning_rate": 8.685125372237374e-05, | |
| "loss": 0.3178, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.7637271214642263, | |
| "grad_norm": 0.07721773788542918, | |
| "learning_rate": 8.646687002498692e-05, | |
| "loss": 0.307, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.767054908485857, | |
| "grad_norm": 0.0804534000899422, | |
| "learning_rate": 8.608268990399349e-05, | |
| "loss": 0.3205, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.7703826955074875, | |
| "grad_norm": 0.0792026473002189, | |
| "learning_rate": 8.569871913854458e-05, | |
| "loss": 0.3172, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.7737104825291181, | |
| "grad_norm": 0.0785093026754779, | |
| "learning_rate": 8.531496350464211e-05, | |
| "loss": 0.3053, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.7770382695507487, | |
| "grad_norm": 0.07651330728326249, | |
| "learning_rate": 8.49314287750517e-05, | |
| "loss": 0.312, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.7803660565723793, | |
| "grad_norm": 0.07697281139664616, | |
| "learning_rate": 8.454812071921596e-05, | |
| "loss": 0.3126, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.78369384359401, | |
| "grad_norm": 0.07825935198806637, | |
| "learning_rate": 8.416504510316773e-05, | |
| "loss": 0.3122, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.7870216306156406, | |
| "grad_norm": 0.07720170469655833, | |
| "learning_rate": 8.378220768944327e-05, | |
| "loss": 0.3176, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.7903494176372712, | |
| "grad_norm": 0.07581499756616941, | |
| "learning_rate": 8.339961423699562e-05, | |
| "loss": 0.3041, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.7936772046589018, | |
| "grad_norm": 0.07874104977091875, | |
| "learning_rate": 8.301727050110793e-05, | |
| "loss": 0.3055, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.7970049916805324, | |
| "grad_norm": 0.07760752760751546, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.3167, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.800332778702163, | |
| "grad_norm": 0.0780397172519416, | |
| "learning_rate": 8.22533551812765e-05, | |
| "loss": 0.317, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.8036605657237936, | |
| "grad_norm": 0.07513849404254262, | |
| "learning_rate": 8.187179508877085e-05, | |
| "loss": 0.3079, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.8069883527454242, | |
| "grad_norm": 0.07993133380201359, | |
| "learning_rate": 8.149050769552856e-05, | |
| "loss": 0.3117, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.8103161397670549, | |
| "grad_norm": 0.0810372072045817, | |
| "learning_rate": 8.1109498737186e-05, | |
| "loss": 0.3111, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.8136439267886857, | |
| "grad_norm": 0.07749863265835616, | |
| "learning_rate": 8.072877394519102e-05, | |
| "loss": 0.2992, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.8169717138103163, | |
| "grad_norm": 0.0798757420555556, | |
| "learning_rate": 8.034833904671698e-05, | |
| "loss": 0.3148, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.820299500831947, | |
| "grad_norm": 0.07823863208458656, | |
| "learning_rate": 7.996819976457626e-05, | |
| "loss": 0.3066, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.8236272878535775, | |
| "grad_norm": 0.07713081216453134, | |
| "learning_rate": 7.958836181713445e-05, | |
| "loss": 0.3125, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.8269550748752081, | |
| "grad_norm": 0.08214863510884096, | |
| "learning_rate": 7.920883091822408e-05, | |
| "loss": 0.306, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.8302828618968388, | |
| "grad_norm": 0.07839949054869876, | |
| "learning_rate": 7.882961277705895e-05, | |
| "loss": 0.3126, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8336106489184694, | |
| "grad_norm": 0.08052815697062654, | |
| "learning_rate": 7.845071309814802e-05, | |
| "loss": 0.3047, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.8369384359401, | |
| "grad_norm": 0.08011461829487185, | |
| "learning_rate": 7.807213758120966e-05, | |
| "loss": 0.3179, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.8402662229617306, | |
| "grad_norm": 0.08011059503152206, | |
| "learning_rate": 7.769389192108608e-05, | |
| "loss": 0.3027, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.8435940099833612, | |
| "grad_norm": 0.07796141187648994, | |
| "learning_rate": 7.731598180765732e-05, | |
| "loss": 0.3218, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.8469217970049918, | |
| "grad_norm": 0.07649200434968564, | |
| "learning_rate": 7.693841292575598e-05, | |
| "loss": 0.304, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.8502495840266224, | |
| "grad_norm": 0.07730271330980557, | |
| "learning_rate": 7.656119095508154e-05, | |
| "loss": 0.3021, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.853577371048253, | |
| "grad_norm": 0.07640278466406251, | |
| "learning_rate": 7.618432157011494e-05, | |
| "loss": 0.3094, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.8569051580698837, | |
| "grad_norm": 0.07947073431868955, | |
| "learning_rate": 7.580781044003324e-05, | |
| "loss": 0.3123, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.8602329450915143, | |
| "grad_norm": 0.0797398351559987, | |
| "learning_rate": 7.543166322862437e-05, | |
| "loss": 0.3218, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.8635607321131449, | |
| "grad_norm": 0.07952419385852674, | |
| "learning_rate": 7.505588559420189e-05, | |
| "loss": 0.3198, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8668885191347755, | |
| "grad_norm": 0.08078121908031825, | |
| "learning_rate": 7.468048318951983e-05, | |
| "loss": 0.3184, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.870216306156406, | |
| "grad_norm": 0.07626793405808584, | |
| "learning_rate": 7.43054616616878e-05, | |
| "loss": 0.3119, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.8735440931780367, | |
| "grad_norm": 0.07724091714553077, | |
| "learning_rate": 7.393082665208586e-05, | |
| "loss": 0.3192, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.8768718801996673, | |
| "grad_norm": 0.07872447703951056, | |
| "learning_rate": 7.35565837962798e-05, | |
| "loss": 0.3039, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.880199667221298, | |
| "grad_norm": 0.07793716133962067, | |
| "learning_rate": 7.318273872393625e-05, | |
| "loss": 0.3127, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.8835274542429286, | |
| "grad_norm": 0.0761903254761463, | |
| "learning_rate": 7.280929705873818e-05, | |
| "loss": 0.3142, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.8868552412645592, | |
| "grad_norm": 0.07585212366951496, | |
| "learning_rate": 7.243626441830009e-05, | |
| "loss": 0.3056, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.8901830282861898, | |
| "grad_norm": 0.07649326568617078, | |
| "learning_rate": 7.206364641408357e-05, | |
| "loss": 0.3016, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.8935108153078204, | |
| "grad_norm": 0.07953013384362162, | |
| "learning_rate": 7.169144865131296e-05, | |
| "loss": 0.3038, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.896838602329451, | |
| "grad_norm": 0.07833536931181964, | |
| "learning_rate": 7.131967672889101e-05, | |
| "loss": 0.313, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.9001663893510816, | |
| "grad_norm": 0.07978547121211514, | |
| "learning_rate": 7.094833623931455e-05, | |
| "loss": 0.308, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.9034941763727122, | |
| "grad_norm": 0.07881716483816299, | |
| "learning_rate": 7.057743276859048e-05, | |
| "loss": 0.3171, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.9068219633943428, | |
| "grad_norm": 0.080242287504338, | |
| "learning_rate": 7.02069718961518e-05, | |
| "loss": 0.3125, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.9101497504159735, | |
| "grad_norm": 0.07778405869750295, | |
| "learning_rate": 6.983695919477345e-05, | |
| "loss": 0.3209, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.913477537437604, | |
| "grad_norm": 0.0780058767785712, | |
| "learning_rate": 6.94674002304887e-05, | |
| "loss": 0.3171, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.9168053244592347, | |
| "grad_norm": 0.081331902124051, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.3227, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.9201331114808653, | |
| "grad_norm": 0.07863423968159455, | |
| "learning_rate": 6.872966574312182e-05, | |
| "loss": 0.3171, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.923460898502496, | |
| "grad_norm": 0.07797145424280164, | |
| "learning_rate": 6.836150131764434e-05, | |
| "loss": 0.3136, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.9267886855241265, | |
| "grad_norm": 0.07683454912238805, | |
| "learning_rate": 6.799381282430284e-05, | |
| "loss": 0.3101, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.9301164725457571, | |
| "grad_norm": 0.07669945961898635, | |
| "learning_rate": 6.762660579416791e-05, | |
| "loss": 0.3058, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9334442595673877, | |
| "grad_norm": 0.07199760768612506, | |
| "learning_rate": 6.725988575106756e-05, | |
| "loss": 0.2962, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.9367720465890184, | |
| "grad_norm": 0.08015294889924407, | |
| "learning_rate": 6.68936582115042e-05, | |
| "loss": 0.3135, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.940099833610649, | |
| "grad_norm": 0.08359235080454425, | |
| "learning_rate": 6.652792868457158e-05, | |
| "loss": 0.3156, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.9434276206322796, | |
| "grad_norm": 0.07912747014264376, | |
| "learning_rate": 6.61627026718719e-05, | |
| "loss": 0.3076, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.9467554076539102, | |
| "grad_norm": 0.07673510899284841, | |
| "learning_rate": 6.579798566743314e-05, | |
| "loss": 0.3053, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.9500831946755408, | |
| "grad_norm": 0.07749051718785945, | |
| "learning_rate": 6.543378315762634e-05, | |
| "loss": 0.3075, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.9534109816971714, | |
| "grad_norm": 0.08129543777852082, | |
| "learning_rate": 6.507010062108309e-05, | |
| "loss": 0.3099, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.956738768718802, | |
| "grad_norm": 0.08006507379003704, | |
| "learning_rate": 6.470694352861312e-05, | |
| "loss": 0.3073, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.9600665557404326, | |
| "grad_norm": 0.08107248449070761, | |
| "learning_rate": 6.4344317343122e-05, | |
| "loss": 0.3158, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.9633943427620633, | |
| "grad_norm": 0.07605456396856061, | |
| "learning_rate": 6.398222751952899e-05, | |
| "loss": 0.3027, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9667221297836939, | |
| "grad_norm": 0.08211872873948112, | |
| "learning_rate": 6.362067950468489e-05, | |
| "loss": 0.3084, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.9700499168053245, | |
| "grad_norm": 0.08166474427359101, | |
| "learning_rate": 6.325967873729018e-05, | |
| "loss": 0.3128, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.973377703826955, | |
| "grad_norm": 0.07877759513153408, | |
| "learning_rate": 6.289923064781332e-05, | |
| "loss": 0.3116, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.9767054908485857, | |
| "grad_norm": 0.07699575903624799, | |
| "learning_rate": 6.25393406584088e-05, | |
| "loss": 0.3002, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.9800332778702163, | |
| "grad_norm": 0.0798025547545203, | |
| "learning_rate": 6.218001418283576e-05, | |
| "loss": 0.3196, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.983361064891847, | |
| "grad_norm": 0.08075862126322861, | |
| "learning_rate": 6.18212566263765e-05, | |
| "loss": 0.3182, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.9866888519134775, | |
| "grad_norm": 0.07976569505261802, | |
| "learning_rate": 6.146307338575519e-05, | |
| "loss": 0.3118, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.9900166389351082, | |
| "grad_norm": 0.07624652784979599, | |
| "learning_rate": 6.110546984905661e-05, | |
| "loss": 0.3061, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.9933444259567388, | |
| "grad_norm": 0.07685433759752576, | |
| "learning_rate": 6.074845139564529e-05, | |
| "loss": 0.3061, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.9966722129783694, | |
| "grad_norm": 0.07734387969968742, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.309, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.07694640533126676, | |
| "learning_rate": 6.0036191212054795e-05, | |
| "loss": 0.3012, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.3221759796142578, | |
| "eval_runtime": 70.8991, | |
| "eval_samples_per_second": 28.562, | |
| "eval_steps_per_second": 0.903, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.0033277870216306, | |
| "grad_norm": 0.07523718675196417, | |
| "learning_rate": 5.9680960196274994e-05, | |
| "loss": 0.2921, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.0066555740432612, | |
| "grad_norm": 0.0744986198393667, | |
| "learning_rate": 5.9326335692419995e-05, | |
| "loss": 0.2955, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.009983361064892, | |
| "grad_norm": 0.07782529390305652, | |
| "learning_rate": 5.89723230350412e-05, | |
| "loss": 0.2964, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.0133111480865225, | |
| "grad_norm": 0.07610051820018607, | |
| "learning_rate": 5.861892754948609e-05, | |
| "loss": 0.2916, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.016638935108153, | |
| "grad_norm": 0.07655912820707372, | |
| "learning_rate": 5.8266154551818216e-05, | |
| "loss": 0.2819, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.0199667221297837, | |
| "grad_norm": 0.07926531609007606, | |
| "learning_rate": 5.79140093487371e-05, | |
| "loss": 0.2931, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.0232945091514143, | |
| "grad_norm": 0.08419237478844939, | |
| "learning_rate": 5.756249723749847e-05, | |
| "loss": 0.3109, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.026622296173045, | |
| "grad_norm": 0.08129019481043527, | |
| "learning_rate": 5.72116235058346e-05, | |
| "loss": 0.2815, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.0299500831946755, | |
| "grad_norm": 0.08162249890462565, | |
| "learning_rate": 5.6861393431874675e-05, | |
| "loss": 0.2912, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.033277870216306, | |
| "grad_norm": 0.0816255564263364, | |
| "learning_rate": 5.651181228406553e-05, | |
| "loss": 0.2874, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.0366056572379367, | |
| "grad_norm": 0.08257234318532701, | |
| "learning_rate": 5.616288532109225e-05, | |
| "loss": 0.2907, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.0399334442595674, | |
| "grad_norm": 0.0813183539547941, | |
| "learning_rate": 5.581461779179924e-05, | |
| "loss": 0.2904, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.043261231281198, | |
| "grad_norm": 0.08205686604462313, | |
| "learning_rate": 5.546701493511106e-05, | |
| "loss": 0.2965, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.0465890183028286, | |
| "grad_norm": 0.08153972009520385, | |
| "learning_rate": 5.5120081979953785e-05, | |
| "loss": 0.288, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.049916805324459, | |
| "grad_norm": 0.08232744842041952, | |
| "learning_rate": 5.477382414517624e-05, | |
| "loss": 0.2857, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.05324459234609, | |
| "grad_norm": 0.08638852212210564, | |
| "learning_rate": 5.442824663947157e-05, | |
| "loss": 0.3014, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.0565723793677204, | |
| "grad_norm": 0.08082652699068096, | |
| "learning_rate": 5.4083354661298814e-05, | |
| "loss": 0.2955, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.059900166389351, | |
| "grad_norm": 0.08036355621896225, | |
| "learning_rate": 5.373915339880484e-05, | |
| "loss": 0.2943, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.0632279534109816, | |
| "grad_norm": 0.07922666094737961, | |
| "learning_rate": 5.339564802974615e-05, | |
| "loss": 0.2839, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0665557404326123, | |
| "grad_norm": 0.08270520249320817, | |
| "learning_rate": 5.305284372141095e-05, | |
| "loss": 0.2989, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.069883527454243, | |
| "grad_norm": 0.08382926768503525, | |
| "learning_rate": 5.2710745630541666e-05, | |
| "loss": 0.3077, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.0732113144758735, | |
| "grad_norm": 0.08125137045700305, | |
| "learning_rate": 5.236935890325717e-05, | |
| "loss": 0.2955, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.076539101497504, | |
| "grad_norm": 0.0850748348065721, | |
| "learning_rate": 5.2028688674975415e-05, | |
| "loss": 0.3, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.0798668885191347, | |
| "grad_norm": 0.08093321024813238, | |
| "learning_rate": 5.168874007033615e-05, | |
| "loss": 0.2897, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.0831946755407653, | |
| "grad_norm": 0.08201475028886479, | |
| "learning_rate": 5.134951820312401e-05, | |
| "loss": 0.2972, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.086522462562396, | |
| "grad_norm": 0.08489256641708065, | |
| "learning_rate": 5.101102817619131e-05, | |
| "loss": 0.2974, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.0898502495840265, | |
| "grad_norm": 0.08171580446986279, | |
| "learning_rate": 5.0673275081381475e-05, | |
| "loss": 0.2908, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.093178036605657, | |
| "grad_norm": 0.08264469368611166, | |
| "learning_rate": 5.0336263999452406e-05, | |
| "loss": 0.2852, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.0965058236272878, | |
| "grad_norm": 0.08334141370794539, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.2962, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0998336106489184, | |
| "grad_norm": 0.08213568371655847, | |
| "learning_rate": 4.966448814138203e-05, | |
| "loss": 0.2915, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.103161397670549, | |
| "grad_norm": 0.0811810267730468, | |
| "learning_rate": 4.932973347064177e-05, | |
| "loss": 0.2988, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.1064891846921796, | |
| "grad_norm": 0.0821423599440937, | |
| "learning_rate": 4.899574102343247e-05, | |
| "loss": 0.2865, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.10981697171381, | |
| "grad_norm": 0.0818046354851144, | |
| "learning_rate": 4.8662515823941255e-05, | |
| "loss": 0.2926, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.113144758735441, | |
| "grad_norm": 0.0797578583724167, | |
| "learning_rate": 4.833006288481371e-05, | |
| "loss": 0.2865, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.1164725457570714, | |
| "grad_norm": 0.08497973521892348, | |
| "learning_rate": 4.799838720707846e-05, | |
| "loss": 0.2903, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.119800332778702, | |
| "grad_norm": 0.08301814094341477, | |
| "learning_rate": 4.766749378007193e-05, | |
| "loss": 0.3007, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.1231281198003327, | |
| "grad_norm": 0.08212424260879446, | |
| "learning_rate": 4.733738758136327e-05, | |
| "loss": 0.286, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.1264559068219633, | |
| "grad_norm": 0.08167559215838262, | |
| "learning_rate": 4.700807357667952e-05, | |
| "loss": 0.2994, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.129783693843594, | |
| "grad_norm": 0.08009449999854285, | |
| "learning_rate": 4.66795567198309e-05, | |
| "loss": 0.2895, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.1331114808652245, | |
| "grad_norm": 0.08145984969869143, | |
| "learning_rate": 4.635184195263624e-05, | |
| "loss": 0.2806, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.136439267886855, | |
| "grad_norm": 0.08284001703324301, | |
| "learning_rate": 4.6024934204848745e-05, | |
| "loss": 0.2909, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.1397670549084857, | |
| "grad_norm": 0.08413611493862337, | |
| "learning_rate": 4.56988383940817e-05, | |
| "loss": 0.293, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.1430948419301163, | |
| "grad_norm": 0.08280804692336827, | |
| "learning_rate": 4.537355942573463e-05, | |
| "loss": 0.2865, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.146422628951747, | |
| "grad_norm": 0.08375099988359626, | |
| "learning_rate": 4.50491021929194e-05, | |
| "loss": 0.2914, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.1497504159733776, | |
| "grad_norm": 0.08658478798479996, | |
| "learning_rate": 4.4725471576386735e-05, | |
| "loss": 0.3019, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.153078202995008, | |
| "grad_norm": 0.08272177787109958, | |
| "learning_rate": 4.4402672444452666e-05, | |
| "loss": 0.2954, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.156405990016639, | |
| "grad_norm": 0.08371985074415418, | |
| "learning_rate": 4.4080709652925336e-05, | |
| "loss": 0.2969, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.1597337770382694, | |
| "grad_norm": 0.08288064281429573, | |
| "learning_rate": 4.3759588045032006e-05, | |
| "loss": 0.3046, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.1630615640599, | |
| "grad_norm": 0.0841149265700377, | |
| "learning_rate": 4.343931245134616e-05, | |
| "loss": 0.3007, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.1663893510815306, | |
| "grad_norm": 0.08290664771215894, | |
| "learning_rate": 4.3119887689714844e-05, | |
| "loss": 0.2815, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.1697171381031612, | |
| "grad_norm": 0.0842022428364363, | |
| "learning_rate": 4.2801318565186165e-05, | |
| "loss": 0.2935, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.173044925124792, | |
| "grad_norm": 0.08367148928442285, | |
| "learning_rate": 4.248360986993711e-05, | |
| "loss": 0.3023, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.1763727121464225, | |
| "grad_norm": 0.08285058082286284, | |
| "learning_rate": 4.216676638320135e-05, | |
| "loss": 0.29, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.179700499168053, | |
| "grad_norm": 0.08263412584525412, | |
| "learning_rate": 4.185079287119733e-05, | |
| "loss": 0.3042, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.1830282861896837, | |
| "grad_norm": 0.08631746371121711, | |
| "learning_rate": 4.15356940870567e-05, | |
| "loss": 0.2921, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.1863560732113143, | |
| "grad_norm": 0.08011738213922821, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.2914, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.189683860232945, | |
| "grad_norm": 0.07957350646648807, | |
| "learning_rate": 4.090813964902889e-05, | |
| "loss": 0.2901, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.1930116472545755, | |
| "grad_norm": 0.08332093195517987, | |
| "learning_rate": 4.0595693435328086e-05, | |
| "loss": 0.2953, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.196339434276206, | |
| "grad_norm": 0.08393402606234097, | |
| "learning_rate": 4.028414082972141e-05, | |
| "loss": 0.2893, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1996672212978368, | |
| "grad_norm": 0.08209487762480656, | |
| "learning_rate": 3.9973486518837565e-05, | |
| "loss": 0.283, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.2029950083194674, | |
| "grad_norm": 0.08521014816831722, | |
| "learning_rate": 3.966373517579244e-05, | |
| "loss": 0.2954, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.206322795341098, | |
| "grad_norm": 0.08356749895415638, | |
| "learning_rate": 3.935489146011869e-05, | |
| "loss": 0.2854, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.2096505823627286, | |
| "grad_norm": 0.08384726016866333, | |
| "learning_rate": 3.904696001769571e-05, | |
| "loss": 0.2909, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.212978369384359, | |
| "grad_norm": 0.0810149357346147, | |
| "learning_rate": 3.873994548067972e-05, | |
| "loss": 0.2883, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.21630615640599, | |
| "grad_norm": 0.08526993500744631, | |
| "learning_rate": 3.843385246743417e-05, | |
| "loss": 0.2957, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.2196339434276204, | |
| "grad_norm": 0.08309493318241117, | |
| "learning_rate": 3.8128685582460145e-05, | |
| "loss": 0.2924, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.222961730449251, | |
| "grad_norm": 0.08272955715006117, | |
| "learning_rate": 3.7824449416327126e-05, | |
| "loss": 0.2926, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.2262895174708817, | |
| "grad_norm": 0.08582812237822925, | |
| "learning_rate": 3.7521148545604e-05, | |
| "loss": 0.2907, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.2296173044925123, | |
| "grad_norm": 0.08273893471578908, | |
| "learning_rate": 3.721878753279017e-05, | |
| "loss": 0.2905, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.232945091514143, | |
| "grad_norm": 0.08532196370072426, | |
| "learning_rate": 3.691737092624688e-05, | |
| "loss": 0.2966, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.2362728785357735, | |
| "grad_norm": 0.08368005152875357, | |
| "learning_rate": 3.661690326012897e-05, | |
| "loss": 0.2982, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.239600665557404, | |
| "grad_norm": 0.082180625122743, | |
| "learning_rate": 3.631738905431641e-05, | |
| "loss": 0.2923, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.2429284525790347, | |
| "grad_norm": 0.08423783786004256, | |
| "learning_rate": 3.601883281434652e-05, | |
| "loss": 0.2937, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.2462562396006653, | |
| "grad_norm": 0.08240657690935907, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.298, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.249584026622296, | |
| "grad_norm": 0.08319470727526912, | |
| "learning_rate": 3.542461218196379e-05, | |
| "loss": 0.2955, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.2529118136439266, | |
| "grad_norm": 0.0808540570268943, | |
| "learning_rate": 3.5128956728303e-05, | |
| "loss": 0.2933, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.256239600665557, | |
| "grad_norm": 0.08128768813005571, | |
| "learning_rate": 3.483427711785449e-05, | |
| "loss": 0.2937, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.259567387687188, | |
| "grad_norm": 0.08743566917375886, | |
| "learning_rate": 3.4540577783429626e-05, | |
| "loss": 0.292, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.2628951747088184, | |
| "grad_norm": 0.08512723696041279, | |
| "learning_rate": 3.424786314309365e-05, | |
| "loss": 0.2919, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.266222961730449, | |
| "grad_norm": 0.08467660314219548, | |
| "learning_rate": 3.395613760009925e-05, | |
| "loss": 0.2993, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.2695507487520796, | |
| "grad_norm": 0.08195600260185569, | |
| "learning_rate": 3.366540554282028e-05, | |
| "loss": 0.2883, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.2728785357737102, | |
| "grad_norm": 0.0861648054687708, | |
| "learning_rate": 3.337567134468579e-05, | |
| "loss": 0.3089, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.276206322795341, | |
| "grad_norm": 0.08393337299331617, | |
| "learning_rate": 3.308693936411421e-05, | |
| "loss": 0.2844, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.2795341098169715, | |
| "grad_norm": 0.08284816292259548, | |
| "learning_rate": 3.279921394444776e-05, | |
| "loss": 0.2904, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.2828618968386025, | |
| "grad_norm": 0.08425001329003003, | |
| "learning_rate": 3.2512499413887255e-05, | |
| "loss": 0.2956, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.286189683860233, | |
| "grad_norm": 0.08215462564052614, | |
| "learning_rate": 3.222680008542678e-05, | |
| "loss": 0.2863, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.2895174708818637, | |
| "grad_norm": 0.08491294093589744, | |
| "learning_rate": 3.194212025678896e-05, | |
| "loss": 0.2979, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.2928452579034944, | |
| "grad_norm": 0.08544022810212593, | |
| "learning_rate": 3.1658464210360284e-05, | |
| "loss": 0.3025, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.296173044925125, | |
| "grad_norm": 0.08519230692029282, | |
| "learning_rate": 3.137583621312665e-05, | |
| "loss": 0.2942, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2995008319467556, | |
| "grad_norm": 0.08648688950410402, | |
| "learning_rate": 3.10942405166092e-05, | |
| "loss": 0.2931, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.302828618968386, | |
| "grad_norm": 0.0855686386000237, | |
| "learning_rate": 3.0813681356800405e-05, | |
| "loss": 0.2892, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.306156405990017, | |
| "grad_norm": 0.08298285994543841, | |
| "learning_rate": 3.053416295410026e-05, | |
| "loss": 0.2956, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.3094841930116474, | |
| "grad_norm": 0.08483591598574476, | |
| "learning_rate": 3.025568951325287e-05, | |
| "loss": 0.2869, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.312811980033278, | |
| "grad_norm": 0.08429953782818361, | |
| "learning_rate": 2.997826522328315e-05, | |
| "loss": 0.2965, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.3161397670549086, | |
| "grad_norm": 0.08652082571572639, | |
| "learning_rate": 2.9701894257433826e-05, | |
| "loss": 0.2906, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.3194675540765393, | |
| "grad_norm": 0.08102566174090828, | |
| "learning_rate": 2.9426580773102698e-05, | |
| "loss": 0.2922, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.32279534109817, | |
| "grad_norm": 0.08191377127522201, | |
| "learning_rate": 2.9152328911780026e-05, | |
| "loss": 0.2886, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.3261231281198005, | |
| "grad_norm": 0.08415392353945142, | |
| "learning_rate": 2.8879142798986292e-05, | |
| "loss": 0.297, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.329450915141431, | |
| "grad_norm": 0.08296700248133196, | |
| "learning_rate": 2.8607026544210114e-05, | |
| "loss": 0.2907, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3327787021630617, | |
| "grad_norm": 0.08376381263415836, | |
| "learning_rate": 2.8335984240846426e-05, | |
| "loss": 0.2872, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.3361064891846923, | |
| "grad_norm": 0.08346616638984133, | |
| "learning_rate": 2.8066019966134904e-05, | |
| "loss": 0.2854, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.339434276206323, | |
| "grad_norm": 0.08433324019878777, | |
| "learning_rate": 2.779713778109867e-05, | |
| "loss": 0.2881, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.3427620632279536, | |
| "grad_norm": 0.08360715945864314, | |
| "learning_rate": 2.7529341730483117e-05, | |
| "loss": 0.2984, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.346089850249584, | |
| "grad_norm": 0.08581998838566726, | |
| "learning_rate": 2.7262635842695127e-05, | |
| "loss": 0.2954, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.3494176372712148, | |
| "grad_norm": 0.0825680191537002, | |
| "learning_rate": 2.6997024129742542e-05, | |
| "loss": 0.2965, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.3527454242928454, | |
| "grad_norm": 0.0825230312739672, | |
| "learning_rate": 2.6732510587173643e-05, | |
| "loss": 0.2973, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.356073211314476, | |
| "grad_norm": 0.08455671065834447, | |
| "learning_rate": 2.6469099194017143e-05, | |
| "loss": 0.295, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.3594009983361066, | |
| "grad_norm": 0.08424896917474452, | |
| "learning_rate": 2.620679391272236e-05, | |
| "loss": 0.2951, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.3627287853577372, | |
| "grad_norm": 0.08530485145974072, | |
| "learning_rate": 2.594559868909956e-05, | |
| "loss": 0.2982, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.366056572379368, | |
| "grad_norm": 0.08142730156733052, | |
| "learning_rate": 2.5685517452260567e-05, | |
| "loss": 0.2882, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.3693843594009985, | |
| "grad_norm": 0.08265505240557701, | |
| "learning_rate": 2.542655411455982e-05, | |
| "loss": 0.288, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.372712146422629, | |
| "grad_norm": 0.08541111432550712, | |
| "learning_rate": 2.5168712571535303e-05, | |
| "loss": 0.2911, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.3760399334442597, | |
| "grad_norm": 0.08297255085261755, | |
| "learning_rate": 2.491199670185008e-05, | |
| "loss": 0.2959, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.3793677204658903, | |
| "grad_norm": 0.08314892429660471, | |
| "learning_rate": 2.465641036723393e-05, | |
| "loss": 0.2974, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.382695507487521, | |
| "grad_norm": 0.08684786516409969, | |
| "learning_rate": 2.4401957412425214e-05, | |
| "loss": 0.2968, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.3860232945091515, | |
| "grad_norm": 0.08074981366720223, | |
| "learning_rate": 2.4148641665113113e-05, | |
| "loss": 0.2881, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.389351081530782, | |
| "grad_norm": 0.08622288522550288, | |
| "learning_rate": 2.389646693587996e-05, | |
| "loss": 0.2989, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.3926788685524127, | |
| "grad_norm": 0.0843914629937067, | |
| "learning_rate": 2.3645437018143978e-05, | |
| "loss": 0.2868, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.3960066555740434, | |
| "grad_norm": 0.08470290466125838, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.2996, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.399334442595674, | |
| "grad_norm": 0.08150938235758469, | |
| "learning_rate": 2.3146826704673696e-05, | |
| "loss": 0.2869, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.4026622296173046, | |
| "grad_norm": 0.08320915618548369, | |
| "learning_rate": 2.2899253809442944e-05, | |
| "loss": 0.2873, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.405990016638935, | |
| "grad_norm": 0.08403664582581712, | |
| "learning_rate": 2.265284072660362e-05, | |
| "loss": 0.2908, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.409317803660566, | |
| "grad_norm": 0.08639583864294073, | |
| "learning_rate": 2.2407591162902573e-05, | |
| "loss": 0.2962, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.4126455906821964, | |
| "grad_norm": 0.08240975441989226, | |
| "learning_rate": 2.2163508807583998e-05, | |
| "loss": 0.2837, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.415973377703827, | |
| "grad_norm": 0.08682037934762359, | |
| "learning_rate": 2.192059733233408e-05, | |
| "loss": 0.2942, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.4193011647254576, | |
| "grad_norm": 0.08318389871173766, | |
| "learning_rate": 2.1678860391225586e-05, | |
| "loss": 0.2915, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.4226289517470883, | |
| "grad_norm": 0.0838348023646509, | |
| "learning_rate": 2.1438301620662993e-05, | |
| "loss": 0.2909, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.425956738768719, | |
| "grad_norm": 0.08527364848116248, | |
| "learning_rate": 2.119892463932781e-05, | |
| "loss": 0.2994, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.4292845257903495, | |
| "grad_norm": 0.08283518402283026, | |
| "learning_rate": 2.0960733048124083e-05, | |
| "loss": 0.2849, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.43261231281198, | |
| "grad_norm": 0.08630649685523206, | |
| "learning_rate": 2.0723730430124223e-05, | |
| "loss": 0.3037, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.4359400998336107, | |
| "grad_norm": 0.08269518183846858, | |
| "learning_rate": 2.0487920350515212e-05, | |
| "loss": 0.2909, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.4392678868552413, | |
| "grad_norm": 0.08642985073342663, | |
| "learning_rate": 2.0253306356544842e-05, | |
| "loss": 0.2898, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.442595673876872, | |
| "grad_norm": 0.08404388826523775, | |
| "learning_rate": 2.0019891977468408e-05, | |
| "loss": 0.2927, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.4459234608985025, | |
| "grad_norm": 0.08578695259137233, | |
| "learning_rate": 1.9787680724495617e-05, | |
| "loss": 0.284, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.449251247920133, | |
| "grad_norm": 0.08490405137348037, | |
| "learning_rate": 1.95566760907378e-05, | |
| "loss": 0.2956, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.4525790349417638, | |
| "grad_norm": 0.08441160147102547, | |
| "learning_rate": 1.9326881551155307e-05, | |
| "loss": 0.2888, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.4559068219633944, | |
| "grad_norm": 0.08515977531799197, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.2955, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.459234608985025, | |
| "grad_norm": 0.08249568872612259, | |
| "learning_rate": 1.88709365632896e-05, | |
| "loss": 0.2841, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.4625623960066556, | |
| "grad_norm": 0.08553334731501643, | |
| "learning_rate": 1.864479297370325e-05, | |
| "loss": 0.2948, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.465890183028286, | |
| "grad_norm": 0.08505739229758723, | |
| "learning_rate": 1.8419873195582814e-05, | |
| "loss": 0.2843, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 2.469217970049917, | |
| "grad_norm": 0.0857918969879419, | |
| "learning_rate": 1.819618061235525e-05, | |
| "loss": 0.2889, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 2.4725457570715474, | |
| "grad_norm": 0.08485052938192404, | |
| "learning_rate": 1.7973718588987097e-05, | |
| "loss": 0.2915, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 2.475873544093178, | |
| "grad_norm": 0.08604466832277857, | |
| "learning_rate": 1.775249047193377e-05, | |
| "loss": 0.2886, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 2.4792013311148087, | |
| "grad_norm": 0.08624664019384962, | |
| "learning_rate": 1.7532499589089323e-05, | |
| "loss": 0.2937, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.4825291181364393, | |
| "grad_norm": 0.08704803308809336, | |
| "learning_rate": 1.7313749249736267e-05, | |
| "loss": 0.2884, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 2.48585690515807, | |
| "grad_norm": 0.08631029427208195, | |
| "learning_rate": 1.7096242744495837e-05, | |
| "loss": 0.2948, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 2.4891846921797005, | |
| "grad_norm": 0.07948765268167073, | |
| "learning_rate": 1.687998334527853e-05, | |
| "loss": 0.2777, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 2.492512479201331, | |
| "grad_norm": 0.0857004189702385, | |
| "learning_rate": 1.6664974305234847e-05, | |
| "loss": 0.291, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 2.4958402662229617, | |
| "grad_norm": 0.08630898293400029, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.2971, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.4991680532445923, | |
| "grad_norm": 0.08585670353827018, | |
| "learning_rate": 1.6238720221177063e-05, | |
| "loss": 0.2878, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 2.502495840266223, | |
| "grad_norm": 0.08375384745783325, | |
| "learning_rate": 1.6027481589225026e-05, | |
| "loss": 0.2873, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 2.5058236272878536, | |
| "grad_norm": 0.08480824204774734, | |
| "learning_rate": 1.5817506140474247e-05, | |
| "loss": 0.2899, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 2.509151414309484, | |
| "grad_norm": 0.08272501630514685, | |
| "learning_rate": 1.560879703354693e-05, | |
| "loss": 0.2919, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.512479201331115, | |
| "grad_norm": 0.08431506160438339, | |
| "learning_rate": 1.5401357408015893e-05, | |
| "loss": 0.2926, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.5158069883527454, | |
| "grad_norm": 0.0854712109543807, | |
| "learning_rate": 1.5195190384357404e-05, | |
| "loss": 0.2914, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 2.519134775374376, | |
| "grad_norm": 0.08456238958113006, | |
| "learning_rate": 1.4990299063904201e-05, | |
| "loss": 0.2962, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 2.5224625623960066, | |
| "grad_norm": 0.08451333183625064, | |
| "learning_rate": 1.4786686528798876e-05, | |
| "loss": 0.2952, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 2.5257903494176372, | |
| "grad_norm": 0.08618684801403191, | |
| "learning_rate": 1.458435584194745e-05, | |
| "loss": 0.3074, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 2.529118136439268, | |
| "grad_norm": 0.08333772045133589, | |
| "learning_rate": 1.4383310046973365e-05, | |
| "loss": 0.2892, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5324459234608985, | |
| "grad_norm": 0.08335870581057453, | |
| "learning_rate": 1.4183552168171654e-05, | |
| "loss": 0.2837, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 2.535773710482529, | |
| "grad_norm": 0.08490274778203831, | |
| "learning_rate": 1.3985085210463477e-05, | |
| "loss": 0.2913, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 2.5391014975041597, | |
| "grad_norm": 0.08303260363269974, | |
| "learning_rate": 1.3787912159350902e-05, | |
| "loss": 0.2911, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 2.5424292845257903, | |
| "grad_norm": 0.08604498795880948, | |
| "learning_rate": 1.3592035980871953e-05, | |
| "loss": 0.3028, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 2.545757071547421, | |
| "grad_norm": 0.08750926191333681, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 0.2893, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.5490848585690515, | |
| "grad_norm": 0.08281648728258575, | |
| "learning_rate": 1.3204186008379927e-05, | |
| "loss": 0.2876, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 2.552412645590682, | |
| "grad_norm": 0.08435910686409348, | |
| "learning_rate": 1.3012218048722858e-05, | |
| "loss": 0.2819, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 2.5557404326123128, | |
| "grad_norm": 0.08339607452067774, | |
| "learning_rate": 1.2821558630323772e-05, | |
| "loss": 0.2863, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 2.5590682196339434, | |
| "grad_norm": 0.08379658010947157, | |
| "learning_rate": 1.2632210621237328e-05, | |
| "loss": 0.2913, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 2.562396006655574, | |
| "grad_norm": 0.08401401125379367, | |
| "learning_rate": 1.2444176869790925e-05, | |
| "loss": 0.2875, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.5657237936772046, | |
| "grad_norm": 0.08260322949768045, | |
| "learning_rate": 1.2257460204541794e-05, | |
| "loss": 0.2952, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 2.569051580698835, | |
| "grad_norm": 0.08240904469376385, | |
| "learning_rate": 1.2072063434234559e-05, | |
| "loss": 0.2847, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 2.572379367720466, | |
| "grad_norm": 0.08487498578135273, | |
| "learning_rate": 1.188798934775881e-05, | |
| "loss": 0.3026, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 2.5757071547420964, | |
| "grad_norm": 0.08507270326525296, | |
| "learning_rate": 1.1705240714107302e-05, | |
| "loss": 0.2886, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 2.579034941763727, | |
| "grad_norm": 0.08446091554336568, | |
| "learning_rate": 1.1523820282334219e-05, | |
| "loss": 0.2944, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.5823627287853577, | |
| "grad_norm": 0.08350296889818828, | |
| "learning_rate": 1.1343730781513895e-05, | |
| "loss": 0.2961, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 2.5856905158069883, | |
| "grad_norm": 0.0845089337024481, | |
| "learning_rate": 1.116497492069961e-05, | |
| "loss": 0.2821, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 2.589018302828619, | |
| "grad_norm": 0.08372812296840969, | |
| "learning_rate": 1.0987555388883041e-05, | |
| "loss": 0.2927, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 2.5923460898502495, | |
| "grad_norm": 0.08452698770354679, | |
| "learning_rate": 1.0811474854953707e-05, | |
| "loss": 0.303, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 2.59567387687188, | |
| "grad_norm": 0.08514333726834006, | |
| "learning_rate": 1.0636735967658784e-05, | |
| "loss": 0.2859, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.5990016638935107, | |
| "grad_norm": 0.08635866955069134, | |
| "learning_rate": 1.0463341355563317e-05, | |
| "loss": 0.2969, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 2.6023294509151413, | |
| "grad_norm": 0.08455859472649895, | |
| "learning_rate": 1.029129362701068e-05, | |
| "loss": 0.2972, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 2.605657237936772, | |
| "grad_norm": 0.08364338552006162, | |
| "learning_rate": 1.0120595370083318e-05, | |
| "loss": 0.295, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 2.6089850249584026, | |
| "grad_norm": 0.08289198213636688, | |
| "learning_rate": 9.95124915256378e-06, | |
| "loss": 0.2982, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 2.612312811980033, | |
| "grad_norm": 0.08268035708483505, | |
| "learning_rate": 9.783257521896227e-06, | |
| "loss": 0.2907, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.615640599001664, | |
| "grad_norm": 0.0834360363580441, | |
| "learning_rate": 9.616623005147951e-06, | |
| "loss": 0.2861, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 2.6189683860232944, | |
| "grad_norm": 0.08537504857757641, | |
| "learning_rate": 9.451348108971426e-06, | |
| "loss": 0.2936, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 2.622296173044925, | |
| "grad_norm": 0.08389496014528498, | |
| "learning_rate": 9.287435319566618e-06, | |
| "loss": 0.2906, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 2.6256239600665556, | |
| "grad_norm": 0.08232408797265886, | |
| "learning_rate": 9.124887102643575e-06, | |
| "loss": 0.2917, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 2.6289517470881862, | |
| "grad_norm": 0.08543468820471654, | |
| "learning_rate": 8.963705903385345e-06, | |
| "loss": 0.2849, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.632279534109817, | |
| "grad_norm": 0.08597027740619363, | |
| "learning_rate": 8.803894146411118e-06, | |
| "loss": 0.2825, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 2.6356073211314475, | |
| "grad_norm": 0.08497563119375516, | |
| "learning_rate": 8.645454235739903e-06, | |
| "loss": 0.2887, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 2.638935108153078, | |
| "grad_norm": 0.08585441969917841, | |
| "learning_rate": 8.488388554754222e-06, | |
| "loss": 0.2996, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 2.6422628951747087, | |
| "grad_norm": 0.08256201957452125, | |
| "learning_rate": 8.332699466164306e-06, | |
| "loss": 0.2883, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 2.6455906821963393, | |
| "grad_norm": 0.08497719915602263, | |
| "learning_rate": 8.178389311972612e-06, | |
| "loss": 0.2967, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.64891846921797, | |
| "grad_norm": 0.08402687681775904, | |
| "learning_rate": 8.025460413438457e-06, | |
| "loss": 0.279, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 2.6522462562396005, | |
| "grad_norm": 0.08477562763677908, | |
| "learning_rate": 7.873915071043247e-06, | |
| "loss": 0.2857, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 2.655574043261231, | |
| "grad_norm": 0.08104346670324097, | |
| "learning_rate": 7.72375556445577e-06, | |
| "loss": 0.2901, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 2.6589018302828618, | |
| "grad_norm": 0.08552766317481185, | |
| "learning_rate": 7.574984152497988e-06, | |
| "loss": 0.2925, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 2.6622296173044924, | |
| "grad_norm": 0.08369865616993412, | |
| "learning_rate": 7.427603073110967e-06, | |
| "loss": 0.2845, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.665557404326123, | |
| "grad_norm": 0.08487536368233828, | |
| "learning_rate": 7.281614543321269e-06, | |
| "loss": 0.2809, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 2.6688851913477536, | |
| "grad_norm": 0.08511471967290406, | |
| "learning_rate": 7.13702075920758e-06, | |
| "loss": 0.288, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 2.672212978369384, | |
| "grad_norm": 0.08464576782836698, | |
| "learning_rate": 6.99382389586769e-06, | |
| "loss": 0.2934, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 2.675540765391015, | |
| "grad_norm": 0.08441009035160585, | |
| "learning_rate": 6.852026107385756e-06, | |
| "loss": 0.3033, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 2.6788685524126454, | |
| "grad_norm": 0.08273775082175298, | |
| "learning_rate": 6.7116295267999455e-06, | |
| "loss": 0.2902, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.682196339434276, | |
| "grad_norm": 0.08352638992484965, | |
| "learning_rate": 6.572636266070264e-06, | |
| "loss": 0.2863, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 2.6855241264559067, | |
| "grad_norm": 0.08362821410899854, | |
| "learning_rate": 6.435048416046863e-06, | |
| "loss": 0.2897, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 2.6888519134775377, | |
| "grad_norm": 0.08364909379187364, | |
| "learning_rate": 6.298868046438533e-06, | |
| "loss": 0.2849, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 2.6921797004991683, | |
| "grad_norm": 0.08410020402772843, | |
| "learning_rate": 6.164097205781616e-06, | |
| "loss": 0.2904, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 2.695507487520799, | |
| "grad_norm": 0.0854674987956861, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.2864, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.6988352745424296, | |
| "grad_norm": 0.08460366688368436, | |
| "learning_rate": 5.898792199420445e-06, | |
| "loss": 0.2918, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 2.70216306156406, | |
| "grad_norm": 0.0852054855400647, | |
| "learning_rate": 5.768262024650773e-06, | |
| "loss": 0.2869, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 2.7054908485856908, | |
| "grad_norm": 0.08602669246911199, | |
| "learning_rate": 5.639149360641649e-06, | |
| "loss": 0.2954, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 2.7088186356073214, | |
| "grad_norm": 0.08442069356281844, | |
| "learning_rate": 5.511456149611194e-06, | |
| "loss": 0.2903, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 2.712146422628952, | |
| "grad_norm": 0.0872199784994378, | |
| "learning_rate": 5.385184312424974e-06, | |
| "loss": 0.2919, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.7154742096505826, | |
| "grad_norm": 0.08392935935083369, | |
| "learning_rate": 5.26033574856708e-06, | |
| "loss": 0.2933, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 2.7188019966722132, | |
| "grad_norm": 0.0866324140535497, | |
| "learning_rate": 5.136912336111599e-06, | |
| "loss": 0.3029, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 2.722129783693844, | |
| "grad_norm": 0.0859681390431694, | |
| "learning_rate": 5.014915931694253e-06, | |
| "loss": 0.2872, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 2.7254575707154745, | |
| "grad_norm": 0.08452596417669088, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.2946, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 2.728785357737105, | |
| "grad_norm": 0.08782758749996201, | |
| "learning_rate": 4.775211466158469e-06, | |
| "loss": 0.2765, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7321131447587357, | |
| "grad_norm": 0.08144696465479902, | |
| "learning_rate": 4.657507010870343e-06, | |
| "loss": 0.2981, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 2.7354409317803663, | |
| "grad_norm": 0.08135983560899517, | |
| "learning_rate": 4.541236775226809e-06, | |
| "loss": 0.2908, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 2.738768718801997, | |
| "grad_norm": 0.08409909748594922, | |
| "learning_rate": 4.426402508259709e-06, | |
| "loss": 0.288, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 2.7420965058236275, | |
| "grad_norm": 0.08304067275962851, | |
| "learning_rate": 4.313005937399861e-06, | |
| "loss": 0.2906, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 2.745424292845258, | |
| "grad_norm": 0.08496162164444254, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.2964, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.7487520798668887, | |
| "grad_norm": 0.08240946749443678, | |
| "learning_rate": 4.0905326855646185e-06, | |
| "loss": 0.2907, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 2.7520798668885194, | |
| "grad_norm": 0.0847301098297941, | |
| "learning_rate": 3.981459351213568e-06, | |
| "loss": 0.2927, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 2.75540765391015, | |
| "grad_norm": 0.08290579774816521, | |
| "learning_rate": 3.873830406168111e-06, | |
| "loss": 0.2936, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 2.7587354409317806, | |
| "grad_norm": 0.08160696067215166, | |
| "learning_rate": 3.7676474694707698e-06, | |
| "loss": 0.2855, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 2.762063227953411, | |
| "grad_norm": 0.08277113220978395, | |
| "learning_rate": 3.662912138411967e-06, | |
| "loss": 0.2846, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.765391014975042, | |
| "grad_norm": 0.08549873509616365, | |
| "learning_rate": 3.5596259885061102e-06, | |
| "loss": 0.2944, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 2.7687188019966724, | |
| "grad_norm": 0.0845502332338962, | |
| "learning_rate": 3.457790573467812e-06, | |
| "loss": 0.3028, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 2.772046589018303, | |
| "grad_norm": 0.08426370264308386, | |
| "learning_rate": 3.357407425188541e-06, | |
| "loss": 0.2906, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 2.7753743760399336, | |
| "grad_norm": 0.08392479123746849, | |
| "learning_rate": 3.2584780537136207e-06, | |
| "loss": 0.2866, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 2.7787021630615643, | |
| "grad_norm": 0.08234066094426105, | |
| "learning_rate": 3.161003947219421e-06, | |
| "loss": 0.2883, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.782029950083195, | |
| "grad_norm": 0.08273035793127474, | |
| "learning_rate": 3.0649865719910798e-06, | |
| "loss": 0.2864, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 2.7853577371048255, | |
| "grad_norm": 0.0848470294206127, | |
| "learning_rate": 2.970427372400353e-06, | |
| "loss": 0.2919, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 2.788685524126456, | |
| "grad_norm": 0.0844832120097032, | |
| "learning_rate": 2.877327770883964e-06, | |
| "loss": 0.2925, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 2.7920133111480867, | |
| "grad_norm": 0.08476915124231062, | |
| "learning_rate": 2.7856891679221564e-06, | |
| "loss": 0.295, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 2.7953410981697173, | |
| "grad_norm": 0.08277761617667297, | |
| "learning_rate": 2.6955129420176196e-06, | |
| "loss": 0.2915, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.798668885191348, | |
| "grad_norm": 0.08691095474681947, | |
| "learning_rate": 2.6068004496747956e-06, | |
| "loss": 0.2968, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 2.8019966722129785, | |
| "grad_norm": 0.08553607826034598, | |
| "learning_rate": 2.5195530253794396e-06, | |
| "loss": 0.2929, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 2.805324459234609, | |
| "grad_norm": 0.08371791879951158, | |
| "learning_rate": 2.433771981578581e-06, | |
| "loss": 0.2889, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 2.8086522462562398, | |
| "grad_norm": 0.08529347009830832, | |
| "learning_rate": 2.349458608660704e-06, | |
| "loss": 0.2918, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 2.8119800332778704, | |
| "grad_norm": 0.08476097984032646, | |
| "learning_rate": 2.266614174936443e-06, | |
| "loss": 0.2915, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.815307820299501, | |
| "grad_norm": 0.08550506338616543, | |
| "learning_rate": 2.1852399266194314e-06, | |
| "loss": 0.2979, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 2.8186356073211316, | |
| "grad_norm": 0.08418728119886265, | |
| "learning_rate": 2.1053370878075684e-06, | |
| "loss": 0.2919, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 2.821963394342762, | |
| "grad_norm": 0.08321073026360784, | |
| "learning_rate": 2.026906860464606e-06, | |
| "loss": 0.2839, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 2.825291181364393, | |
| "grad_norm": 0.08521572901725284, | |
| "learning_rate": 1.9499504244020693e-06, | |
| "loss": 0.2895, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 2.8286189683860234, | |
| "grad_norm": 0.08302231632821679, | |
| "learning_rate": 1.874468937261531e-06, | |
| "loss": 0.2856, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.831946755407654, | |
| "grad_norm": 0.08446164481121499, | |
| "learning_rate": 1.8004635344971654e-06, | |
| "loss": 0.2906, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 2.8352745424292847, | |
| "grad_norm": 0.08461614775418828, | |
| "learning_rate": 1.7279353293586765e-06, | |
| "loss": 0.2951, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 2.8386023294509153, | |
| "grad_norm": 0.08376751456819907, | |
| "learning_rate": 1.6568854128745537e-06, | |
| "loss": 0.2867, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 2.841930116472546, | |
| "grad_norm": 0.08485720840208241, | |
| "learning_rate": 1.5873148538356753e-06, | |
| "loss": 0.2902, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 2.8452579034941765, | |
| "grad_norm": 0.08327929136686951, | |
| "learning_rate": 1.5192246987791981e-06, | |
| "loss": 0.2824, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.848585690515807, | |
| "grad_norm": 0.08376005940787881, | |
| "learning_rate": 1.4526159719728594e-06, | |
| "loss": 0.2929, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 2.8519134775374377, | |
| "grad_norm": 0.08576482740883794, | |
| "learning_rate": 1.3874896753995003e-06, | |
| "loss": 0.2883, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 2.8552412645590683, | |
| "grad_norm": 0.08423906037237339, | |
| "learning_rate": 1.323846788742078e-06, | |
| "loss": 0.2892, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 2.858569051580699, | |
| "grad_norm": 0.08555775217042304, | |
| "learning_rate": 1.261688269368877e-06, | |
| "loss": 0.2815, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 2.8618968386023296, | |
| "grad_norm": 0.08352867363337906, | |
| "learning_rate": 1.201015052319099e-06, | |
| "loss": 0.2878, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.86522462562396, | |
| "grad_norm": 0.08043712828269083, | |
| "learning_rate": 1.14182805028884e-06, | |
| "loss": 0.2898, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 2.868552412645591, | |
| "grad_norm": 0.08358283864190634, | |
| "learning_rate": 1.084128153617292e-06, | |
| "loss": 0.2794, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 2.8718801996672214, | |
| "grad_norm": 0.08496321233585508, | |
| "learning_rate": 1.0279162302734624e-06, | |
| "loss": 0.2958, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 2.875207986688852, | |
| "grad_norm": 0.08670014766582711, | |
| "learning_rate": 9.731931258429638e-07, | |
| "loss": 0.2986, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 2.8785357737104826, | |
| "grad_norm": 0.08423391631225062, | |
| "learning_rate": 9.199596635154683e-07, | |
| "loss": 0.2872, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.8818635607321132, | |
| "grad_norm": 0.08213293799990758, | |
| "learning_rate": 8.682166440721728e-07, | |
| "loss": 0.2835, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 2.885191347753744, | |
| "grad_norm": 0.08243067843868014, | |
| "learning_rate": 8.17964845873831e-07, | |
| "loss": 0.2868, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 2.8885191347753745, | |
| "grad_norm": 0.08356376535800905, | |
| "learning_rate": 7.69205024849029e-07, | |
| "loss": 0.2856, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 2.891846921797005, | |
| "grad_norm": 0.08284966743698358, | |
| "learning_rate": 7.219379144828287e-07, | |
| "loss": 0.2905, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 2.8951747088186357, | |
| "grad_norm": 0.08294456984096978, | |
| "learning_rate": 6.761642258056978e-07, | |
| "loss": 0.2818, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.8985024958402663, | |
| "grad_norm": 0.08562316895318861, | |
| "learning_rate": 6.318846473828522e-07, | |
| "loss": 0.2982, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 2.901830282861897, | |
| "grad_norm": 0.08319155744729591, | |
| "learning_rate": 5.890998453038644e-07, | |
| "loss": 0.2871, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 2.9051580698835275, | |
| "grad_norm": 0.08284573396497398, | |
| "learning_rate": 5.478104631726711e-07, | |
| "loss": 0.2875, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 2.908485856905158, | |
| "grad_norm": 0.08238254352208946, | |
| "learning_rate": 5.080171220978813e-07, | |
| "loss": 0.2997, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 2.9118136439267888, | |
| "grad_norm": 0.08315333334198248, | |
| "learning_rate": 4.6972042068341714e-07, | |
| "loss": 0.2881, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.9151414309484194, | |
| "grad_norm": 0.08267169693953713, | |
| "learning_rate": 4.329209350195651e-07, | |
| "loss": 0.2857, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 2.91846921797005, | |
| "grad_norm": 0.08430946661727712, | |
| "learning_rate": 3.976192186742167e-07, | |
| "loss": 0.2945, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 2.9217970049916806, | |
| "grad_norm": 0.0825306082723947, | |
| "learning_rate": 3.638158026846306e-07, | |
| "loss": 0.2892, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 2.925124792013311, | |
| "grad_norm": 0.08616315933414796, | |
| "learning_rate": 3.315111955493944e-07, | |
| "loss": 0.2883, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 2.928452579034942, | |
| "grad_norm": 0.08564491699314337, | |
| "learning_rate": 3.007058832207976e-07, | |
| "loss": 0.2987, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.9317803660565724, | |
| "grad_norm": 0.08213284459782745, | |
| "learning_rate": 2.7140032909749316e-07, | |
| "loss": 0.2932, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 2.935108153078203, | |
| "grad_norm": 0.08599861031926373, | |
| "learning_rate": 2.4359497401758024e-07, | |
| "loss": 0.2908, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 2.9384359400998337, | |
| "grad_norm": 0.08387269862429567, | |
| "learning_rate": 2.1729023625189915e-07, | |
| "loss": 0.2905, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 2.9417637271214643, | |
| "grad_norm": 0.0820198351623876, | |
| "learning_rate": 1.924865114978025e-07, | |
| "loss": 0.2896, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 2.945091514143095, | |
| "grad_norm": 0.08529310320299995, | |
| "learning_rate": 1.6918417287318245e-07, | |
| "loss": 0.2852, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.9484193011647255, | |
| "grad_norm": 0.08351749593178831, | |
| "learning_rate": 1.4738357091084176e-07, | |
| "loss": 0.2828, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 2.951747088186356, | |
| "grad_norm": 0.08388738488232517, | |
| "learning_rate": 1.2708503355323143e-07, | |
| "loss": 0.2914, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 2.9550748752079867, | |
| "grad_norm": 0.08738900028103073, | |
| "learning_rate": 1.0828886614754341e-07, | |
| "loss": 0.3117, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 2.9584026622296173, | |
| "grad_norm": 0.08256854242946138, | |
| "learning_rate": 9.099535144108107e-08, | |
| "loss": 0.2902, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 2.961730449251248, | |
| "grad_norm": 0.08458250620395145, | |
| "learning_rate": 7.520474957699586e-08, | |
| "loss": 0.2818, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9650582362728786, | |
| "grad_norm": 0.08352851941481279, | |
| "learning_rate": 6.09172980904238e-08, | |
| "loss": 0.2889, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 2.968386023294509, | |
| "grad_norm": 0.08267252929606801, | |
| "learning_rate": 4.8133211904888285e-08, | |
| "loss": 0.2881, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.97171381031614, | |
| "grad_norm": 0.08356961669951495, | |
| "learning_rate": 3.685268332905834e-08, | |
| "loss": 0.2898, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 2.9750415973377704, | |
| "grad_norm": 0.08041089699968551, | |
| "learning_rate": 2.7075882053828605e-08, | |
| "loss": 0.2774, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.978369384359401, | |
| "grad_norm": 0.08241694313728697, | |
| "learning_rate": 1.8802955149865852e-08, | |
| "loss": 0.2897, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.9816971713810316, | |
| "grad_norm": 0.08324205581056916, | |
| "learning_rate": 1.2034027065255249e-08, | |
| "loss": 0.2841, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 2.9850249584026622, | |
| "grad_norm": 0.08182160853929105, | |
| "learning_rate": 6.769199623779532e-09, | |
| "loss": 0.2873, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 2.988352745424293, | |
| "grad_norm": 0.08394433913606542, | |
| "learning_rate": 3.0085520232425722e-09, | |
| "loss": 0.2832, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 2.9916805324459235, | |
| "grad_norm": 0.081226371854989, | |
| "learning_rate": 7.521408343924563e-10, | |
| "loss": 0.281, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 2.995008319467554, | |
| "grad_norm": 0.08252475778296225, | |
| "learning_rate": 0.0, | |
| "loss": 0.2901, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.995008319467554, | |
| "eval_loss": 0.3194945156574249, | |
| "eval_runtime": 70.0427, | |
| "eval_samples_per_second": 28.911, | |
| "eval_steps_per_second": 0.914, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.995008319467554, | |
| "step": 900, | |
| "total_flos": 3.649557175842898e+17, | |
| "train_loss": 0.3413675943348143, | |
| "train_runtime": 11378.8153, | |
| "train_samples_per_second": 10.138, | |
| "train_steps_per_second": 0.079 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.649557175842898e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |