| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.051152203381160645, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00010230440676232128, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.751, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00020460881352464257, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.8722, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0003069132202869639, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 3e-06, | |
| "loss": 1.5552, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00040921762704928513, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.8481, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0005115220338116064, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 5e-06, | |
| "loss": 1.6085, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0006138264405739278, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 6e-06, | |
| "loss": 1.8854, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0007161308473362491, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 1.5885, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0008184352540985703, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6188, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0009207396608608916, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 9e-06, | |
| "loss": 1.7513, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0010230440676232129, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 1e-05, | |
| "loss": 1.847, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.001125348474385534, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.5958, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0012276528811478555, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.7284, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0013299572879101767, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.6829, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0014322616946724981, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.8158, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0015345661014348193, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.9778, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0016368705081971405, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.7056, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.001739174914959462, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 1.8864, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0018414793217217831, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.7853, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0019437837284841043, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.7343, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0020460881352464258, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9046, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.002148392542008747, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 2.1e-05, | |
| "loss": 1.8652, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.002250696948771068, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.916, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0023530013555333896, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 1.6101, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.002455305762295711, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.7165, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.002557610169058032, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.9596, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0026599145758203534, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 2.0135, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.002762218982582675, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 1.8613, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0028645233893449962, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.6937, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0029668277961073172, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 2.9e-05, | |
| "loss": 1.8989, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0030691322028696386, | |
| "grad_norm": 0.625, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9366, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.00317143660963196, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 3.1e-05, | |
| "loss": 1.8176, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.003273741016394281, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.965, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0033760454231566025, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.3e-05, | |
| "loss": 1.8058, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.003478349829918924, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.8037, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.003580654236681245, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.9695, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0036829586434435663, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.7983, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0037852630502058877, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 3.7e-05, | |
| "loss": 1.8376, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0038875674569682087, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.6842, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0039898718637305305, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.57, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0040921762704928515, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 4e-05, | |
| "loss": 1.8307, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0041944806772551725, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.6629, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.004296785084017494, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.801, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.004399089490779815, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 4.3e-05, | |
| "loss": 1.7171, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.004501393897542136, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.7683, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.004603698304304458, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.6179, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.004706002711066779, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.8107, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0048083071178291, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.9287, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.004910611524591422, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.8884, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.005012915931353743, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.6786, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.005115220338116064, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 5e-05, | |
| "loss": 1.848, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005217524744878386, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 1.6701, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.005319829151640707, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.7677, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.005422133558403028, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 1.5958, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.00552443796516535, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.794, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.005626742371927671, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.7429, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0057290467786899925, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.898, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0058313511854523135, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 1.9033, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0059336555922146345, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.7433, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.006035959998976956, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 5.9e-05, | |
| "loss": 1.7133, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.006138264405739277, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 6e-05, | |
| "loss": 1.8123, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.006240568812501598, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 6.1e-05, | |
| "loss": 1.604, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.00634287321926392, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.7869, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.006445177626026241, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 6.3e-05, | |
| "loss": 1.559, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.006547482032788562, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.7037, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.006649786439550884, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.7831, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.006752090846313205, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.6768, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.006854395253075526, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 6.7e-05, | |
| "loss": 1.8762, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.006956699659837848, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.8037, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.007059004066600169, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 6.9e-05, | |
| "loss": 1.9943, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.00716130847336249, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 7e-05, | |
| "loss": 1.6346, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.007263612880124812, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 7.1e-05, | |
| "loss": 1.7755, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.007365917286887133, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.9162, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0074682216936494536, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 7.3e-05, | |
| "loss": 1.6629, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.007570526100411775, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.7918, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.007672830507174096, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.8483, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.007775134913936417, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.5213, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.007877439320698738, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 7.7e-05, | |
| "loss": 1.7234, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.007979743727461061, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.5943, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.008082048134223382, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 1.7327, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.008184352540985703, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 8e-05, | |
| "loss": 1.6263, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.008286656947748024, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.6922, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.008388961354510345, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.6937, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.008491265761272666, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 8.3e-05, | |
| "loss": 1.5954, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.008593570168034989, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.7655, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.00869587457479731, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.8177, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.00879817898155963, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.6066, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.008900483388321952, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 8.7e-05, | |
| "loss": 1.7382, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.009002787795084273, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.5273, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.009105092201846595, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 1.633, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.009207396608608916, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.009309701015371237, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.1e-05, | |
| "loss": 1.6444, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.009412005422133558, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.6237, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.00951430982889588, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 1.5545, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.0096166142356582, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.6549, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.009718918642420523, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.5817, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.009821223049182844, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.617, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.009923527455945165, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.7116, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.010025831862707486, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.6373, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.010128136269469807, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.5059, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.010230440676232128, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4517, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01033274508299445, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.999497487437187e-05, | |
| "loss": 1.5179, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.010435049489756772, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.998994974874373e-05, | |
| "loss": 1.4234, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.010537353896519093, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.998492462311558e-05, | |
| "loss": 1.7412, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.010639658303281414, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.997989949748744e-05, | |
| "loss": 1.7474, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.010741962710043735, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 9.99748743718593e-05, | |
| "loss": 1.7978, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.010844267116806056, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.996984924623116e-05, | |
| "loss": 1.5856, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.010946571523568378, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 9.996482412060301e-05, | |
| "loss": 1.6459, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0110488759303307, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.995979899497487e-05, | |
| "loss": 2.6578, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.01115118033709302, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 9.995477386934674e-05, | |
| "loss": 1.5426, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.011253484743855341, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.99497487437186e-05, | |
| "loss": 1.5997, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.011355789150617662, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 9.994472361809045e-05, | |
| "loss": 1.6825, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.011458093557379985, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 9.993969849246232e-05, | |
| "loss": 1.5822, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.011560397964142306, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 9.993467336683417e-05, | |
| "loss": 1.6862, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.011662702370904627, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.992964824120603e-05, | |
| "loss": 1.4022, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.011765006777666948, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 9.99246231155779e-05, | |
| "loss": 1.5491, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.011867311184429269, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.991959798994976e-05, | |
| "loss": 1.5846, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.01196961559119159, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.99145728643216e-05, | |
| "loss": 1.5171, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.012071919997953913, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.990954773869348e-05, | |
| "loss": 1.506, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.012174224404716234, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.990452261306533e-05, | |
| "loss": 1.7183, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.012276528811478555, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.989949748743719e-05, | |
| "loss": 1.6749, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.012378833218240876, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.989447236180905e-05, | |
| "loss": 1.6602, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.012481137625003197, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.988944723618092e-05, | |
| "loss": 1.6384, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.012583442031765518, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.988442211055276e-05, | |
| "loss": 1.6169, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.01268574643852784, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.987939698492463e-05, | |
| "loss": 1.5463, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.012788050845290161, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.987437185929649e-05, | |
| "loss": 1.6794, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.012890355252052482, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.986934673366835e-05, | |
| "loss": 1.5976, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.012992659658814803, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.98643216080402e-05, | |
| "loss": 1.5735, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.013094964065577124, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.985929648241207e-05, | |
| "loss": 1.4446, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.013197268472339445, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.985427135678392e-05, | |
| "loss": 1.5231, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.013299572879101768, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.984924623115578e-05, | |
| "loss": 1.6299, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.013401877285864089, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.984422110552765e-05, | |
| "loss": 1.6611, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.01350418169262641, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.983919597989951e-05, | |
| "loss": 1.6093, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.01360648609938873, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 9.983417085427136e-05, | |
| "loss": 1.5908, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.013708790506151052, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.982914572864322e-05, | |
| "loss": 1.5976, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.013811094912913375, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.982412060301508e-05, | |
| "loss": 1.4695, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.013913399319675696, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.981909547738694e-05, | |
| "loss": 1.4971, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.014015703726438017, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.98140703517588e-05, | |
| "loss": 1.6059, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.014118008133200338, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.980904522613065e-05, | |
| "loss": 1.5415, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.014220312539962658, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 9.980402010050252e-05, | |
| "loss": 1.6073, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.01432261694672498, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.979899497487438e-05, | |
| "loss": 1.4738, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.014424921353487302, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.979396984924624e-05, | |
| "loss": 1.6089, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.014527225760249623, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.978894472361809e-05, | |
| "loss": 1.5486, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.014629530167011944, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.978391959798995e-05, | |
| "loss": 1.5261, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.014731834573774265, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.977889447236181e-05, | |
| "loss": 1.726, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.014834138980536586, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.977386934673367e-05, | |
| "loss": 1.7616, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.014936443387298907, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.976884422110552e-05, | |
| "loss": 1.8157, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.01503874779406123, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.97638190954774e-05, | |
| "loss": 1.5536, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.01514105220082355, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.975879396984925e-05, | |
| "loss": 1.6118, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.015243356607585872, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.975376884422111e-05, | |
| "loss": 1.7143, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.015345661014348193, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.974874371859297e-05, | |
| "loss": 1.4682, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.015447965421110514, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.974371859296483e-05, | |
| "loss": 1.5663, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.015550269827872835, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.973869346733668e-05, | |
| "loss": 1.6771, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.015652574234635157, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.973366834170856e-05, | |
| "loss": 1.6855, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.015754878641397477, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.97286432160804e-05, | |
| "loss": 1.7027, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0158571830481598, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.972361809045227e-05, | |
| "loss": 1.4748, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.015959487454922122, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.971859296482412e-05, | |
| "loss": 1.5864, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.01606179186168444, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.971356783919599e-05, | |
| "loss": 1.5928, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.016164096268446764, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.970854271356784e-05, | |
| "loss": 1.7351, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.016266400675209083, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.97035175879397e-05, | |
| "loss": 1.7008, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.016368705081971406, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.969849246231156e-05, | |
| "loss": 1.4992, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.01647100948873373, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.969346733668343e-05, | |
| "loss": 1.635, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.016573313895496048, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.968844221105527e-05, | |
| "loss": 1.5781, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.01667561830225837, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.968341708542715e-05, | |
| "loss": 1.4841, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.01677792270902069, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.9678391959799e-05, | |
| "loss": 1.3593, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.016880227115783013, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.967336683417086e-05, | |
| "loss": 1.5774, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.016982531522545332, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.966834170854272e-05, | |
| "loss": 1.5977, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.017084835929307655, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.966331658291458e-05, | |
| "loss": 1.4401, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.017187140336069977, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.965829145728643e-05, | |
| "loss": 1.4486, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.017289444742832297, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.96532663316583e-05, | |
| "loss": 1.5923, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.01739174914959462, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.964824120603016e-05, | |
| "loss": 1.6353, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.01749405355635694, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.964321608040202e-05, | |
| "loss": 1.4871, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.01759635796311926, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.963819095477387e-05, | |
| "loss": 1.6157, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.017698662369881584, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.963316582914573e-05, | |
| "loss": 1.5152, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.017800966776643903, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.962814070351759e-05, | |
| "loss": 1.5647, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.017903271183406226, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.962311557788945e-05, | |
| "loss": 1.6144, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.018005575590168545, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.961809045226132e-05, | |
| "loss": 1.5505, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.018107879996930868, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.961306532663316e-05, | |
| "loss": 1.4726, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.01821018440369319, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.960804020100503e-05, | |
| "loss": 1.484, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.01831248881045551, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.960301507537689e-05, | |
| "loss": 1.3879, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.018414793217217833, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.959798994974875e-05, | |
| "loss": 1.5664, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.018517097623980152, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.95929648241206e-05, | |
| "loss": 1.5777, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.018619402030742475, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.958793969849247e-05, | |
| "loss": 1.4148, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.018721706437504794, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.958291457286432e-05, | |
| "loss": 1.5065, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.018824010844267117, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.957788944723619e-05, | |
| "loss": 1.5637, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.01892631525102944, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.957286432160805e-05, | |
| "loss": 1.4061, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.01902861965779176, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.956783919597991e-05, | |
| "loss": 1.5302, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.01913092406455408, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.956281407035176e-05, | |
| "loss": 1.5825, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.0192332284713164, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.955778894472362e-05, | |
| "loss": 1.4345, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.019335532878078723, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.955276381909548e-05, | |
| "loss": 1.4496, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.019437837284841046, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.954773869346734e-05, | |
| "loss": 1.5227, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.019540141691603365, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.954271356783919e-05, | |
| "loss": 1.5828, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.019642446098365688, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.953768844221107e-05, | |
| "loss": 1.4511, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.019744750505128007, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.953266331658292e-05, | |
| "loss": 1.4965, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.01984705491189033, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.952763819095478e-05, | |
| "loss": 1.6729, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.01994935931865265, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.952261306532664e-05, | |
| "loss": 1.7777, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.020051663725414972, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.95175879396985e-05, | |
| "loss": 1.4149, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.020153968132177295, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.951256281407035e-05, | |
| "loss": 1.7133, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.020256272538939614, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.950753768844223e-05, | |
| "loss": 1.5531, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.020358576945701937, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.950251256281408e-05, | |
| "loss": 1.6573, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.020460881352464256, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.949748743718594e-05, | |
| "loss": 1.3231, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02056318575922658, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.94924623115578e-05, | |
| "loss": 1.7199, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.0206654901659889, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.948743718592966e-05, | |
| "loss": 1.4348, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.02076779457275122, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.948241206030151e-05, | |
| "loss": 1.51, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.020870098979513543, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.947738693467337e-05, | |
| "loss": 1.4171, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.020972403386275863, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.947236180904523e-05, | |
| "loss": 1.5366, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.021074707793038185, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.94673366834171e-05, | |
| "loss": 1.5415, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.021177012199800508, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.946231155778894e-05, | |
| "loss": 1.5641, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.021279316606562827, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.94572864321608e-05, | |
| "loss": 1.5054, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.02138162101332515, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.945226130653267e-05, | |
| "loss": 1.5638, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.02148392542008747, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.944723618090453e-05, | |
| "loss": 1.4641, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.021586229826849792, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.944221105527639e-05, | |
| "loss": 1.5075, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.02168853423361211, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.943718592964824e-05, | |
| "loss": 1.3774, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.021790838640374434, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.94321608040201e-05, | |
| "loss": 1.5137, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.021893143047136757, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.942713567839197e-05, | |
| "loss": 1.463, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.021995447453899076, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.942211055276383e-05, | |
| "loss": 1.5722, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.0220977518606614, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.941708542713568e-05, | |
| "loss": 1.5695, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.022200056267423718, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.941206030150754e-05, | |
| "loss": 1.4522, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.02230236067418604, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.94070351758794e-05, | |
| "loss": 1.5649, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.022404665080948363, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.940201005025126e-05, | |
| "loss": 1.3819, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.022506969487710683, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.939698492462311e-05, | |
| "loss": 1.4209, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.022609273894473005, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.939195979899499e-05, | |
| "loss": 1.5313, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.022711578301235325, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.938693467336683e-05, | |
| "loss": 1.6329, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.022813882707997647, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.93819095477387e-05, | |
| "loss": 1.5602, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.02291618711475997, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.937688442211056e-05, | |
| "loss": 1.4092, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.02301849152152229, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.937185929648242e-05, | |
| "loss": 1.507, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.023120795928284612, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.936683417085427e-05, | |
| "loss": 1.3996, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.02322310033504693, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.936180904522614e-05, | |
| "loss": 1.3146, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.023325404741809254, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.935678391959799e-05, | |
| "loss": 1.4642, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.023427709148571573, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.935175879396985e-05, | |
| "loss": 1.5534, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.023530013555333896, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.934673366834172e-05, | |
| "loss": 1.5299, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.02363231796209622, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.934170854271358e-05, | |
| "loss": 1.6027, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.023734622368858538, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.933668341708543e-05, | |
| "loss": 1.3886, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.02383692677562086, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.933165829145729e-05, | |
| "loss": 1.513, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.02393923118238318, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.932663316582915e-05, | |
| "loss": 1.5952, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.024041535589145503, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.932160804020101e-05, | |
| "loss": 1.4902, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.024143839995907825, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.931658291457286e-05, | |
| "loss": 1.5195, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.024246144402670144, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.931155778894474e-05, | |
| "loss": 1.5003, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.024348448809432467, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.930653266331659e-05, | |
| "loss": 1.6125, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.024450753216194786, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.930150753768845e-05, | |
| "loss": 1.4048, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.02455305762295711, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.929648241206031e-05, | |
| "loss": 1.404, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.024655362029719432, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.929145728643217e-05, | |
| "loss": 1.5336, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.02475766643648175, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.928643216080402e-05, | |
| "loss": 1.5893, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.024859970843244074, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.928140703517588e-05, | |
| "loss": 1.4552, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.024962275250006393, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.927638190954774e-05, | |
| "loss": 1.5381, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.025064579656768716, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.927135678391961e-05, | |
| "loss": 1.4568, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.025166884063531035, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.926633165829147e-05, | |
| "loss": 1.3433, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.025269188470293358, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.926130653266332e-05, | |
| "loss": 1.4223, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.02537149287705568, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.925628140703518e-05, | |
| "loss": 1.6224, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.025473797283818, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.925125628140703e-05, | |
| "loss": 1.4886, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.025576101690580322, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.92462311557789e-05, | |
| "loss": 1.3823, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02567840609734264, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.924120603015075e-05, | |
| "loss": 1.6174, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.025780710504104964, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.923618090452261e-05, | |
| "loss": 1.6667, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.025883014910867287, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.923115577889448e-05, | |
| "loss": 1.5771, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.025985319317629606, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.922613065326634e-05, | |
| "loss": 1.3376, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.02608762372439193, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.922110552763819e-05, | |
| "loss": 1.621, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.02618992813115425, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.921608040201006e-05, | |
| "loss": 1.413, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.02629223253791657, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.921105527638191e-05, | |
| "loss": 1.6237, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.02639453694467889, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.920603015075377e-05, | |
| "loss": 1.459, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.026496841351441213, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.920100502512563e-05, | |
| "loss": 1.5026, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.026599145758203536, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.91959798994975e-05, | |
| "loss": 1.517, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.026701450164965855, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.919095477386935e-05, | |
| "loss": 1.6779, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.026803754571728178, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.918592964824122e-05, | |
| "loss": 1.4558, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.026906058978490497, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.918090452261307e-05, | |
| "loss": 1.451, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.02700836338525282, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.917587939698493e-05, | |
| "loss": 1.538, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.027110667792015142, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.917085427135678e-05, | |
| "loss": 1.4029, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.02721297219877746, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.916582914572866e-05, | |
| "loss": 1.6634, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.027315276605539784, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.91608040201005e-05, | |
| "loss": 1.4518, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.027417581012302104, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.915577889447237e-05, | |
| "loss": 1.5737, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.027519885419064426, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.915075376884423e-05, | |
| "loss": 1.4141, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.02762218982582675, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.914572864321609e-05, | |
| "loss": 1.6944, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.02772449423258907, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.914070351758794e-05, | |
| "loss": 1.6906, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.02782679863935139, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.913567839195981e-05, | |
| "loss": 1.7094, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.02792910304611371, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.913065326633166e-05, | |
| "loss": 1.5747, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.028031407452876033, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.912562814070352e-05, | |
| "loss": 1.508, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.028133711859638352, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.912060301507539e-05, | |
| "loss": 1.4204, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.028236016266400675, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.911557788944725e-05, | |
| "loss": 1.3767, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.028338320673162998, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.91105527638191e-05, | |
| "loss": 1.5665, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.028440625079925317, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.910552763819096e-05, | |
| "loss": 1.4584, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.02854292948668764, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.910050251256282e-05, | |
| "loss": 1.5575, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.02864523389344996, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.909547738693468e-05, | |
| "loss": 1.6704, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02874753830021228, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.909045226130653e-05, | |
| "loss": 1.6166, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.028849842706974604, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.90854271356784e-05, | |
| "loss": 1.5372, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.028952147113736924, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.908040201005026e-05, | |
| "loss": 1.5092, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.029054451520499246, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.90753768844221e-05, | |
| "loss": 1.5686, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.029156755927261566, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.907035175879398e-05, | |
| "loss": 1.4031, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.02925906033402389, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.906532663316583e-05, | |
| "loss": 1.4016, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.02936136474078621, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.906030150753769e-05, | |
| "loss": 1.4951, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.02946366914754853, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.905527638190955e-05, | |
| "loss": 1.4838, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.029565973554310853, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.905025125628141e-05, | |
| "loss": 1.6449, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.029668277961073172, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.904522613065326e-05, | |
| "loss": 1.6668, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.029770582367835495, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.904020100502514e-05, | |
| "loss": 1.4133, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.029872886774597814, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.903517587939699e-05, | |
| "loss": 1.5274, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.029975191181360137, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.903015075376885e-05, | |
| "loss": 1.4883, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.03007749558812246, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.902512562814071e-05, | |
| "loss": 1.5292, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.03017979999488478, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.902010050251257e-05, | |
| "loss": 1.5329, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.0302821044016471, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.901507537688442e-05, | |
| "loss": 1.5716, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.03038440880840942, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.901005025125628e-05, | |
| "loss": 1.6152, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.030486713215171744, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.900502512562815e-05, | |
| "loss": 1.5215, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.030589017621934066, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.4974, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.030691322028696386, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.899497487437186e-05, | |
| "loss": 1.429, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03079362643545871, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.898994974874373e-05, | |
| "loss": 1.6325, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.030895930842221028, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.898492462311558e-05, | |
| "loss": 1.3455, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.03099823524898335, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.897989949748744e-05, | |
| "loss": 1.3542, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.03110053965574567, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.89748743718593e-05, | |
| "loss": 1.6288, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.031202844062507992, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.896984924623117e-05, | |
| "loss": 1.567, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.031305148469270315, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.896482412060301e-05, | |
| "loss": 1.4221, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.03140745287603264, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.895979899497489e-05, | |
| "loss": 1.6211, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.031509757282794953, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.895477386934674e-05, | |
| "loss": 1.5028, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.031612061689557276, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.89497487437186e-05, | |
| "loss": 1.4722, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.0317143660963196, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.894472361809046e-05, | |
| "loss": 1.4819, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.03181667050308192, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.893969849246232e-05, | |
| "loss": 1.4361, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.031918974909844244, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.893467336683417e-05, | |
| "loss": 1.3847, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.03202127931660656, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.892964824120604e-05, | |
| "loss": 1.6572, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.03212358372336888, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.89246231155779e-05, | |
| "loss": 1.4203, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.032225888130131206, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.891959798994975e-05, | |
| "loss": 1.5494, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.03232819253689353, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.891457286432161e-05, | |
| "loss": 1.535, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.03243049694365585, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.890954773869347e-05, | |
| "loss": 1.4293, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.03253280135041817, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.890452261306533e-05, | |
| "loss": 1.5293, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.03263510575718049, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.889949748743718e-05, | |
| "loss": 1.321, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.03273741016394281, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.889447236180906e-05, | |
| "loss": 1.3901, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.032839714570705135, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.88894472361809e-05, | |
| "loss": 1.6575, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.03294201897746746, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.888442211055277e-05, | |
| "loss": 1.3796, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.03304432338422977, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.887939698492463e-05, | |
| "loss": 1.5548, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.033146627790992096, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.887437185929649e-05, | |
| "loss": 1.385, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.03324893219775442, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.886934673366834e-05, | |
| "loss": 1.5527, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.03335123660451674, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.886432160804021e-05, | |
| "loss": 1.546, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.03345354101127906, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.885929648241206e-05, | |
| "loss": 1.4091, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.03355584541804138, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.885427135678393e-05, | |
| "loss": 1.4189, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.0336581498248037, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.884924623115577e-05, | |
| "loss": 1.5757, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.033760454231566026, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.884422110552765e-05, | |
| "loss": 1.4861, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.03386275863832835, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.88391959798995e-05, | |
| "loss": 1.4099, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.033965063045090664, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.883417085427136e-05, | |
| "loss": 1.4004, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.03406736745185299, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.882914572864322e-05, | |
| "loss": 1.5382, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.03416967185861531, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.882412060301508e-05, | |
| "loss": 1.5474, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.03427197626537763, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.881909547738693e-05, | |
| "loss": 2.0222, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.034374280672139955, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.881407035175881e-05, | |
| "loss": 1.506, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.03447658507890227, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.880904522613066e-05, | |
| "loss": 1.5479, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.03457888948566459, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.880402010050252e-05, | |
| "loss": 1.4794, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.034681193892426916, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.879899497487438e-05, | |
| "loss": 1.4704, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.03478349829918924, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.879396984924624e-05, | |
| "loss": 1.72, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.03488580270595156, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.878894472361809e-05, | |
| "loss": 1.3659, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.03498810711271388, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.878391959798995e-05, | |
| "loss": 1.5252, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.0350904115194762, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.877889447236182e-05, | |
| "loss": 1.6494, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.03519271592623852, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.877386934673368e-05, | |
| "loss": 1.4853, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.035295020333000846, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.876884422110553e-05, | |
| "loss": 1.4591, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.03539732473976317, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.87638190954774e-05, | |
| "loss": 1.462, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.035499629146525484, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.875879396984925e-05, | |
| "loss": 1.5346, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.03560193355328781, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.875376884422111e-05, | |
| "loss": 1.4419, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.03570423796005013, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.874874371859297e-05, | |
| "loss": 1.5451, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.03580654236681245, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.874371859296482e-05, | |
| "loss": 1.5113, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.035908846773574775, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.873869346733668e-05, | |
| "loss": 1.4416, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.03601115118033709, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.873366834170855e-05, | |
| "loss": 1.6185, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.03611345558709941, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.872864321608041e-05, | |
| "loss": 1.4294, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.036215759993861736, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.872361809045226e-05, | |
| "loss": 1.6404, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.03631806440062406, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.871859296482413e-05, | |
| "loss": 1.3696, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.03642036880738638, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.871356783919598e-05, | |
| "loss": 1.411, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.0365226732141487, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.870854271356784e-05, | |
| "loss": 1.4582, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.03662497762091102, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.870351758793969e-05, | |
| "loss": 1.7792, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.03672728202767334, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.869849246231157e-05, | |
| "loss": 1.517, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.036829586434435665, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.869346733668342e-05, | |
| "loss": 1.3991, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.03693189084119798, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.868844221105528e-05, | |
| "loss": 1.5767, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.037034195247960304, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.868341708542714e-05, | |
| "loss": 1.4882, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.03713649965472263, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.8678391959799e-05, | |
| "loss": 1.5225, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.03723880406148495, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.867336683417085e-05, | |
| "loss": 1.4675, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.03734110846824727, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.866834170854273e-05, | |
| "loss": 1.5146, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.03744341287500959, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.866331658291457e-05, | |
| "loss": 1.4159, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.03754571728177191, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.865829145728644e-05, | |
| "loss": 1.4227, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.03764802168853423, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.86532663316583e-05, | |
| "loss": 1.548, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.037750326095296556, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.864824120603016e-05, | |
| "loss": 1.5389, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.03785263050205888, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.864321608040201e-05, | |
| "loss": 1.4892, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.037954934908821195, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.863819095477388e-05, | |
| "loss": 1.4655, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.03805723931558352, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.863316582914573e-05, | |
| "loss": 1.4212, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.03815954372234584, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.86281407035176e-05, | |
| "loss": 1.5425, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.03826184812910816, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.862311557788944e-05, | |
| "loss": 1.4689, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.038364152535870485, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.861809045226132e-05, | |
| "loss": 1.432, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.0384664569426328, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.861306532663317e-05, | |
| "loss": 1.5188, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.038568761349395124, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.860804020100503e-05, | |
| "loss": 1.3553, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.03867106575615745, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.860301507537689e-05, | |
| "loss": 1.33, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.03877337016291977, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.859798994974875e-05, | |
| "loss": 1.5601, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.03887567456968209, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.85929648241206e-05, | |
| "loss": 1.4254, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.03897797897644441, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.858793969849246e-05, | |
| "loss": 1.5918, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.03908028338320673, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.858291457286433e-05, | |
| "loss": 1.4162, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.03918258778996905, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.857788944723619e-05, | |
| "loss": 1.5642, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.039284892196731376, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.857286432160805e-05, | |
| "loss": 1.4412, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.0393871966034937, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.85678391959799e-05, | |
| "loss": 1.3841, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.039489501010256015, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.856281407035176e-05, | |
| "loss": 1.498, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.03959180541701834, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.855778894472362e-05, | |
| "loss": 1.4214, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.03969410982378066, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.855276381909548e-05, | |
| "loss": 1.4923, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.03979641423054298, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.854773869346733e-05, | |
| "loss": 1.4241, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.0398987186373053, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.85427135678392e-05, | |
| "loss": 1.7142, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.04000102304406762, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.853768844221106e-05, | |
| "loss": 1.6048, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.040103327450829944, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.853266331658292e-05, | |
| "loss": 1.2976, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.04020563185759227, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.852763819095477e-05, | |
| "loss": 1.4336, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.04030793626435459, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 9.852261306532664e-05, | |
| "loss": 1.526, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.040410240671116905, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.851758793969849e-05, | |
| "loss": 1.5614, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.04051254507787923, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.851256281407035e-05, | |
| "loss": 1.5674, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.04061484948464155, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 9.850753768844222e-05, | |
| "loss": 1.4084, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.04071715389140387, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.850251256281408e-05, | |
| "loss": 1.4501, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.040819458298166196, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.849748743718593e-05, | |
| "loss": 1.5337, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.04092176270492851, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.84924623115578e-05, | |
| "loss": 1.4503, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.041024067111690835, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.848743718592965e-05, | |
| "loss": 1.44, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.04112637151845316, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.848241206030151e-05, | |
| "loss": 1.5709, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.04122867592521548, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.847738693467337e-05, | |
| "loss": 1.5478, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.0413309803319778, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.847236180904524e-05, | |
| "loss": 1.2712, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.04143328473874012, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.846733668341709e-05, | |
| "loss": 1.5433, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.04153558914550244, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.846231155778895e-05, | |
| "loss": 1.4632, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.041637893552264764, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.845728643216081e-05, | |
| "loss": 1.5674, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.04174019795902709, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.845226130653267e-05, | |
| "loss": 1.3531, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.04184250236578941, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.844723618090452e-05, | |
| "loss": 1.6524, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.041944806772551725, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.84422110552764e-05, | |
| "loss": 1.4676, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.04204711117931405, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.843718592964824e-05, | |
| "loss": 1.4209, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.04214941558607637, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 9.84321608040201e-05, | |
| "loss": 1.5188, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.04225171999283869, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.842713567839197e-05, | |
| "loss": 1.4647, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.042354024399601016, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.842211055276383e-05, | |
| "loss": 1.3745, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.04245632880636333, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.841708542713568e-05, | |
| "loss": 1.4112, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.042558633213125655, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 9.841206030150754e-05, | |
| "loss": 1.2667, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.04266093761988798, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.84070351758794e-05, | |
| "loss": 1.5598, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.0427632420266503, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.840201005025126e-05, | |
| "loss": 1.4135, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.04286554643341262, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.839698492462313e-05, | |
| "loss": 1.507, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.04296785084017494, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.839195979899497e-05, | |
| "loss": 1.4718, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.04307015524693726, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.838693467336684e-05, | |
| "loss": 1.4224, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.043172459653699584, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.83819095477387e-05, | |
| "loss": 1.4815, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.04327476406046191, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.837688442211056e-05, | |
| "loss": 1.5333, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.04337706846722422, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.837185929648241e-05, | |
| "loss": 1.4247, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.043479372873986545, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.836683417085427e-05, | |
| "loss": 1.5218, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.04358167728074887, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.836180904522613e-05, | |
| "loss": 1.6057, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.04368398168751119, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.8356783919598e-05, | |
| "loss": 1.6459, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.04378628609427351, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.835175879396984e-05, | |
| "loss": 1.3835, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.04388859050103583, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.834673366834172e-05, | |
| "loss": 1.4054, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.04399089490779815, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.834170854271357e-05, | |
| "loss": 1.4183, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.044093199314560474, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.833668341708543e-05, | |
| "loss": 1.509, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.0441955037213228, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.833165829145729e-05, | |
| "loss": 1.5587, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.04429780812808512, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.832663316582915e-05, | |
| "loss": 1.6455, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.044400112534847436, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.8321608040201e-05, | |
| "loss": 1.3427, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.04450241694160976, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.831658291457288e-05, | |
| "loss": 1.382, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.04460472134837208, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.831155778894473e-05, | |
| "loss": 1.42, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.044707025755134404, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.830653266331659e-05, | |
| "loss": 1.4817, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.04480933016189673, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.830150753768844e-05, | |
| "loss": 1.5843, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.04491163456865904, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.829648241206031e-05, | |
| "loss": 1.4559, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.045013938975421365, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.829145728643216e-05, | |
| "loss": 1.4828, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.04511624338218369, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.828643216080402e-05, | |
| "loss": 1.7695, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.04521854778894601, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.828140703517589e-05, | |
| "loss": 1.3626, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.04532085219570833, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.827638190954775e-05, | |
| "loss": 1.3116, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.04542315660247065, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.82713567839196e-05, | |
| "loss": 1.3595, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.04552546100923297, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.826633165829147e-05, | |
| "loss": 1.5105, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.045627765415995294, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.826130653266332e-05, | |
| "loss": 1.5128, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.04573006982275762, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.825628140703518e-05, | |
| "loss": 1.3877, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.04583237422951994, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.825125628140704e-05, | |
| "loss": 1.5253, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.045934678636282256, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.82462311557789e-05, | |
| "loss": 1.3744, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.04603698304304458, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.824120603015075e-05, | |
| "loss": 1.4214, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0461392874498069, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.823618090452262e-05, | |
| "loss": 1.3398, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.046241591856569224, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.823115577889448e-05, | |
| "loss": 1.5119, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.04634389626333154, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.822613065326634e-05, | |
| "loss": 1.5934, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.04644620067009386, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.822110552763819e-05, | |
| "loss": 1.4602, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.046548505076856185, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.821608040201005e-05, | |
| "loss": 1.5107, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.04665080948361851, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.821105527638191e-05, | |
| "loss": 1.6103, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.04675311389038083, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.820603015075378e-05, | |
| "loss": 1.5265, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.046855418297143146, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.820100502512564e-05, | |
| "loss": 1.3611, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.04695772270390547, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.819597989949749e-05, | |
| "loss": 1.6182, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.04706002711066779, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.819095477386935e-05, | |
| "loss": 1.3604, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.047162331517430114, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.818592964824121e-05, | |
| "loss": 1.5613, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.04726463592419244, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.818090452261307e-05, | |
| "loss": 1.5404, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.04736694033095475, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.817587939698492e-05, | |
| "loss": 1.5799, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.047469244737717076, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.81708542713568e-05, | |
| "loss": 1.2581, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.0475715491444794, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.816582914572864e-05, | |
| "loss": 1.4294, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.04767385355124172, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 9.81608040201005e-05, | |
| "loss": 1.4916, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.047776157958004044, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.815577889447236e-05, | |
| "loss": 1.3478, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.04787846236476636, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.815075376884423e-05, | |
| "loss": 1.5497, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.04798076677152868, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.814572864321608e-05, | |
| "loss": 1.5409, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.048083071178291005, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.814070351758794e-05, | |
| "loss": 1.3988, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.04818537558505333, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.81356783919598e-05, | |
| "loss": 2.4223, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.04828767999181565, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.813065326633167e-05, | |
| "loss": 1.3415, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.048389984398577966, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.812562814070351e-05, | |
| "loss": 1.3554, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.04849228880534029, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.812060301507539e-05, | |
| "loss": 1.5009, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.04859459321210261, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.811557788944724e-05, | |
| "loss": 1.6927, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.048696897618864934, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.81105527638191e-05, | |
| "loss": 1.6283, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.04879920202562726, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.810552763819096e-05, | |
| "loss": 1.434, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.04890150643238957, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.810050251256282e-05, | |
| "loss": 1.522, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.049003810839151896, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.809547738693467e-05, | |
| "loss": 1.4241, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.04910611524591422, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.809045226130655e-05, | |
| "loss": 1.4464, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.04920841965267654, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.80854271356784e-05, | |
| "loss": 1.5481, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.049310724059438864, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.808040201005026e-05, | |
| "loss": 1.6201, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.04941302846620118, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.807537688442211e-05, | |
| "loss": 1.4617, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.0495153328729635, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.807035175879398e-05, | |
| "loss": 1.5015, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.049617637279725825, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 9.806532663316583e-05, | |
| "loss": 1.5654, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.04971994168648815, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.806030150753769e-05, | |
| "loss": 1.4838, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.049822246093250463, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.805527638190956e-05, | |
| "loss": 1.4796, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.049924550500012786, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.805025125628142e-05, | |
| "loss": 1.5272, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.05002685490677511, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.804522613065327e-05, | |
| "loss": 1.6112, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.05012915931353743, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.804020100502513e-05, | |
| "loss": 1.4579, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.050231463720299754, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.803517587939699e-05, | |
| "loss": 1.4171, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.05033376812706207, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.803015075376885e-05, | |
| "loss": 1.6056, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.05043607253382439, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.802512562814071e-05, | |
| "loss": 1.5128, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.050538376940586716, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.802010050251256e-05, | |
| "loss": 1.55, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.05064068134734904, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.801507537688442e-05, | |
| "loss": 1.5617, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.05074298575411136, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.801005025125629e-05, | |
| "loss": 1.4943, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.05084529016087368, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.800502512562815e-05, | |
| "loss": 1.5396, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.050947594567636, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.5213, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.05104989897439832, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.799497487437186e-05, | |
| "loss": 1.3655, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.051152203381160645, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.798994974874372e-05, | |
| "loss": 1.4502, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20000, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.29157249892352e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |