| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.961937716262976, | |
| "eval_steps": 500, | |
| "global_step": 432, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.625, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 1.3219, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 1.2583, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9e-07, | |
| "loss": 1.3233, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.3118, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.75, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.2718, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.625, | |
| "learning_rate": 1.8e-06, | |
| "loss": 1.2916, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.1e-06, | |
| "loss": 1.3006, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.125, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.2396, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 2.7e-06, | |
| "loss": 1.225, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 3e-06, | |
| "loss": 1.2836, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 2.999976893879632e-06, | |
| "loss": 1.2149, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.0, | |
| "learning_rate": 2.999907576230383e-06, | |
| "loss": 1.173, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 2.999792049187804e-06, | |
| "loss": 1.1601, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.375, | |
| "learning_rate": 2.99963031631107e-06, | |
| "loss": 1.1468, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.0, | |
| "learning_rate": 2.9994223825828736e-06, | |
| "loss": 1.1263, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 2.9991682544092705e-06, | |
| "loss": 1.1591, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 2.9988679396194814e-06, | |
| "loss": 1.0788, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.9985214474656536e-06, | |
| "loss": 1.1426, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.375, | |
| "learning_rate": 2.9981287886225726e-06, | |
| "loss": 1.0888, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 3.25, | |
| "learning_rate": 2.997689975187335e-06, | |
| "loss": 1.083, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.75, | |
| "learning_rate": 2.997205020678976e-06, | |
| "loss": 1.071, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.75, | |
| "learning_rate": 2.9966739400380512e-06, | |
| "loss": 1.0907, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.9960967496261784e-06, | |
| "loss": 1.072, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.5, | |
| "learning_rate": 2.995473467225533e-06, | |
| "loss": 1.0573, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.9948041120382984e-06, | |
| "loss": 1.0487, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 2.994088704686077e-06, | |
| "loss": 1.0156, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 2.993327267209254e-06, | |
| "loss": 1.0971, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 2.992519823066316e-06, | |
| "loss": 1.0575, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.375, | |
| "learning_rate": 2.991666397133133e-06, | |
| "loss": 1.028, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 2.9907670157021875e-06, | |
| "loss": 1.0419, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.625, | |
| "learning_rate": 2.9898217064817673e-06, | |
| "loss": 1.15, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.98883049859511e-06, | |
| "loss": 1.0245, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 2.987793422579508e-06, | |
| "loss": 1.0391, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.9867105103853648e-06, | |
| "loss": 1.0343, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.985581795375214e-06, | |
| "loss": 1.0316, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.984407312322688e-06, | |
| "loss": 1.0453, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 2.98318709741145e-06, | |
| "loss": 1.0455, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.9819211882340754e-06, | |
| "loss": 1.033, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 2.9806096237908986e-06, | |
| "loss": 1.0303, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.9792524444888073e-06, | |
| "loss": 1.0006, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 2.977849692139999e-06, | |
| "loss": 1.029, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.25, | |
| "learning_rate": 2.976401409960693e-06, | |
| "loss": 1.048, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.9749076425697996e-06, | |
| "loss": 1.0164, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.973368435987544e-06, | |
| "loss": 1.0075, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.9717838376340497e-06, | |
| "loss": 1.0364, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.9701538963278767e-06, | |
| "loss": 1.0127, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.968478662284519e-06, | |
| "loss": 0.9968, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.9667581871148553e-06, | |
| "loss": 1.0087, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.9649925238235613e-06, | |
| "loss": 0.9926, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.963181726807475e-06, | |
| "loss": 1.0005, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.9613258518539204e-06, | |
| "loss": 1.0029, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.9594249561389925e-06, | |
| "loss": 1.0475, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.957479098225791e-06, | |
| "loss": 0.9926, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.9554883380626183e-06, | |
| "loss": 1.0115, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.9534527369811333e-06, | |
| "loss": 0.9977, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.951372357694461e-06, | |
| "loss": 1.0862, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.94924726429526e-06, | |
| "loss": 1.0044, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.947077522253749e-06, | |
| "loss": 0.9761, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.9448631984156895e-06, | |
| "loss": 1.0414, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.9426043610003245e-06, | |
| "loss": 0.9801, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.9403010795982812e-06, | |
| "loss": 0.9832, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.9379534251694214e-06, | |
| "loss": 0.9952, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.93556147004066e-06, | |
| "loss": 0.9998, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.933125287903734e-06, | |
| "loss": 0.9714, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.9306449538129346e-06, | |
| "loss": 0.9968, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.9281205441827918e-06, | |
| "loss": 1.0091, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.925552136785723e-06, | |
| "loss": 0.9731, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.9229398107496354e-06, | |
| "loss": 0.9712, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.920283646555489e-06, | |
| "loss": 0.9796, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.917583726034816e-06, | |
| "loss": 1.0048, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.9148401323672014e-06, | |
| "loss": 0.9751, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.9120529500777193e-06, | |
| "loss": 1.0171, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.9092222650343288e-06, | |
| "loss": 1.019, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.90634816444523e-06, | |
| "loss": 0.9982, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.903430736856174e-06, | |
| "loss": 0.9658, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.900470072147738e-06, | |
| "loss": 0.9983, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.8974662615325564e-06, | |
| "loss": 0.9966, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.894419397552508e-06, | |
| "loss": 0.9998, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8913295740758676e-06, | |
| "loss": 0.9919, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.888196886294413e-06, | |
| "loss": 1.0071, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.885021430720492e-06, | |
| "loss": 0.9911, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.8818033051840494e-06, | |
| "loss": 1.0248, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.878542608829613e-06, | |
| "loss": 0.9623, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.87523944211324e-06, | |
| "loss": 0.9953, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.87189390679942e-06, | |
| "loss": 0.9831, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.8685061059579428e-06, | |
| "loss": 0.9882, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.8650761439607194e-06, | |
| "loss": 1.0026, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8616041264785704e-06, | |
| "loss": 0.9555, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8580901604779672e-06, | |
| "loss": 0.9886, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.854534354217738e-06, | |
| "loss": 0.9699, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8509368172457333e-06, | |
| "loss": 0.9705, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.8472976603954494e-06, | |
| "loss": 0.9835, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.843616995782614e-06, | |
| "loss": 0.9595, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.8398949368017344e-06, | |
| "loss": 1.0608, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.8361315981226e-06, | |
| "loss": 1.0303, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.832327095686753e-06, | |
| "loss": 0.9669, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.828481546703915e-06, | |
| "loss": 1.0021, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.8245950696483764e-06, | |
| "loss": 0.972, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.8206677842553465e-06, | |
| "loss": 0.9734, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8166998115172635e-06, | |
| "loss": 0.9715, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.812691273680068e-06, | |
| "loss": 0.9896, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.808642294239438e-06, | |
| "loss": 0.9473, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.8045529979369805e-06, | |
| "loss": 0.9784, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.800423510756394e-06, | |
| "loss": 0.9566, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.796253959919581e-06, | |
| "loss": 0.9875, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.7920444738827332e-06, | |
| "loss": 0.9869, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.7877951823323727e-06, | |
| "loss": 0.9945, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.783506216181354e-06, | |
| "loss": 0.9652, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.7791777075648364e-06, | |
| "loss": 0.9878, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.7748097898362074e-06, | |
| "loss": 0.9595, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.770402597562977e-06, | |
| "loss": 0.9677, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.765956266522632e-06, | |
| "loss": 0.9791, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.761470933698453e-06, | |
| "loss": 0.9541, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.7569467372752936e-06, | |
| "loss": 0.9573, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.752383816635323e-06, | |
| "loss": 0.9416, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.747782312353733e-06, | |
| "loss": 0.9575, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.7431423661944057e-06, | |
| "loss": 0.9932, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.7384641211055474e-06, | |
| "loss": 0.9394, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.733747721215283e-06, | |
| "loss": 0.9775, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.728993311827218e-06, | |
| "loss": 0.9711, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.724201039415959e-06, | |
| "loss": 0.9576, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.7193710516226047e-06, | |
| "loss": 0.9695, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.714503497250193e-06, | |
| "loss": 0.979, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.709598526259121e-06, | |
| "loss": 0.9854, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.7046562897625218e-06, | |
| "loss": 0.9882, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.699676940021611e-06, | |
| "loss": 0.9738, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.694660630440994e-06, | |
| "loss": 0.9692, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.6896075155639417e-06, | |
| "loss": 0.9882, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.684517751067629e-06, | |
| "loss": 0.9721, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.6793914937583356e-06, | |
| "loss": 0.9631, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.6742289015666204e-06, | |
| "loss": 0.9752, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.6690301335424527e-06, | |
| "loss": 0.9938, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.663795349850312e-06, | |
| "loss": 0.9598, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.6585247117642533e-06, | |
| "loss": 0.9711, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.6532183816629417e-06, | |
| "loss": 0.9913, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.6478765230246463e-06, | |
| "loss": 0.9741, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.6424993004222054e-06, | |
| "loss": 0.9541, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.637086879517956e-06, | |
| "loss": 0.9721, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.6316394270586294e-06, | |
| "loss": 0.9622, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.6261571108702162e-06, | |
| "loss": 0.9486, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.620640099852793e-06, | |
| "loss": 0.9579, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.6150885639753198e-06, | |
| "loss": 0.9682, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.6095026742704063e-06, | |
| "loss": 1.0175, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.6038826028290377e-06, | |
| "loss": 0.9619, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.598228522795278e-06, | |
| "loss": 1.0443, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.5925406083609323e-06, | |
| "loss": 0.9441, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.5868190347601805e-06, | |
| "loss": 1.0083, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.5810639782641804e-06, | |
| "loss": 0.9488, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 2.0, | |
| "learning_rate": 2.575275616175637e-06, | |
| "loss": 0.9293, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.5694541268233384e-06, | |
| "loss": 0.9648, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.563599689556662e-06, | |
| "loss": 0.9557, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.557712484740051e-06, | |
| "loss": 0.9478, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.551792693747457e-06, | |
| "loss": 0.9569, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.5458404989567504e-06, | |
| "loss": 1.0341, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.5398560837441037e-06, | |
| "loss": 0.9795, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.5338396324783407e-06, | |
| "loss": 0.9526, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.527791330515258e-06, | |
| "loss": 0.9531, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.5217113641919136e-06, | |
| "loss": 0.9193, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.5155999208208857e-06, | |
| "loss": 0.949, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.5094571886845013e-06, | |
| "loss": 0.9595, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.5032833570290396e-06, | |
| "loss": 0.916, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.4970786160588956e-06, | |
| "loss": 0.9826, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.4908431569307268e-06, | |
| "loss": 1.0362, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.4845771717475563e-06, | |
| "loss": 0.9425, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.478280853552863e-06, | |
| "loss": 0.9522, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.4719543963246275e-06, | |
| "loss": 0.9401, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.4655979949693605e-06, | |
| "loss": 0.9662, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.459211845316094e-06, | |
| "loss": 0.9211, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.4527961441103515e-06, | |
| "loss": 0.9594, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.4463510890080865e-06, | |
| "loss": 0.9591, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.4398768785695893e-06, | |
| "loss": 0.92, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.433373712253376e-06, | |
| "loss": 0.9829, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.426841790410037e-06, | |
| "loss": 0.961, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.4202813142760685e-06, | |
| "loss": 0.9428, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.413692485967673e-06, | |
| "loss": 0.9671, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.40707550847453e-06, | |
| "loss": 0.9428, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.400430585653544e-06, | |
| "loss": 0.9587, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.3937579222225646e-06, | |
| "loss": 0.9817, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.387057723754076e-06, | |
| "loss": 0.9479, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.38033019666887e-06, | |
| "loss": 0.9572, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.3735755482296788e-06, | |
| "loss": 0.9411, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.3667939865347966e-06, | |
| "loss": 0.9485, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.359985720511665e-06, | |
| "loss": 0.9688, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.353150959910435e-06, | |
| "loss": 0.9546, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.346289915297509e-06, | |
| "loss": 0.9551, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.33940279804905e-06, | |
| "loss": 0.9427, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.332489820344472e-06, | |
| "loss": 0.9466, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.325551195159902e-06, | |
| "loss": 0.9604, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.3185871362616168e-06, | |
| "loss": 0.9481, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.311597858199462e-06, | |
| "loss": 0.9491, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.3045835763002377e-06, | |
| "loss": 0.9384, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.2975445066610655e-06, | |
| "loss": 0.9591, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.290480866142733e-06, | |
| "loss": 0.9196, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.2833928723630116e-06, | |
| "loss": 0.956, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.2762807436899505e-06, | |
| "loss": 0.9479, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.269144699235152e-06, | |
| "loss": 0.9414, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.2619849588470177e-06, | |
| "loss": 0.9403, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.2548017431039793e-06, | |
| "loss": 0.9684, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.247595273307701e-06, | |
| "loss": 0.9428, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.240365771476259e-06, | |
| "loss": 0.9678, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.233113460337308e-06, | |
| "loss": 0.9463, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.2258385633212116e-06, | |
| "loss": 0.929, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.218541304554167e-06, | |
| "loss": 1.0058, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.2112219088512914e-06, | |
| "loss": 0.9762, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.2038806017097022e-06, | |
| "loss": 0.9186, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.1965176093015694e-06, | |
| "loss": 0.9142, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.1891331584671447e-06, | |
| "loss": 0.9081, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.1817274767077735e-06, | |
| "loss": 0.9452, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.1743007921788887e-06, | |
| "loss": 0.9451, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.166853333682979e-06, | |
| "loss": 0.9489, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.159385330662541e-06, | |
| "loss": 0.9469, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.1518970131930106e-06, | |
| "loss": 0.9583, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.144388611975674e-06, | |
| "loss": 0.9624, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.1368603583305603e-06, | |
| "loss": 0.9407, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.129312484189317e-06, | |
| "loss": 0.9319, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.1217452220880612e-06, | |
| "loss": 0.9579, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 2.1141588051602192e-06, | |
| "loss": 0.9314, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.1065534671293404e-06, | |
| "loss": 0.9589, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.0, | |
| "learning_rate": 2.0989294423018997e-06, | |
| "loss": 0.9262, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.0912869655600775e-06, | |
| "loss": 0.9575, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.0836262723545242e-06, | |
| "loss": 0.9536, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.0759475986971055e-06, | |
| "loss": 0.9357, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.0682511811536306e-06, | |
| "loss": 0.9544, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.0605372568365683e-06, | |
| "loss": 0.9383, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.052806063397736e-06, | |
| "loss": 0.9486, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.045057839020981e-06, | |
| "loss": 0.9822, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.0372928224148454e-06, | |
| "loss": 0.9422, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.029511252805205e-06, | |
| "loss": 0.9402, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.0217133699279074e-06, | |
| "loss": 0.9067, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 2.013899414021378e-06, | |
| "loss": 0.9232, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.006069625819225e-06, | |
| "loss": 0.9292, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9982242465428196e-06, | |
| "loss": 0.9309, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.990363517893867e-06, | |
| "loss": 0.9446, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.9824876820469553e-06, | |
| "loss": 0.9676, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.9745969816421e-06, | |
| "loss": 0.9603, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9666916597772663e-06, | |
| "loss": 0.9393, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9587719600008777e-06, | |
| "loss": 0.9352, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9508381263043165e-06, | |
| "loss": 0.9233, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.942890403114404e-06, | |
| "loss": 0.9573, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9349290352858735e-06, | |
| "loss": 0.9531, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9269542680938204e-06, | |
| "loss": 0.9406, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.918966347226154e-06, | |
| "loss": 0.934, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.910965518776022e-06, | |
| "loss": 0.9359, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.9029520292342306e-06, | |
| "loss": 0.9453, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.8949261254816526e-06, | |
| "loss": 0.9516, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.8868880547816187e-06, | |
| "loss": 0.9814, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.8788380647723008e-06, | |
| "loss": 0.9549, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.870776403459083e-06, | |
| "loss": 0.9348, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.8627033192069213e-06, | |
| "loss": 0.9241, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.8546190607326902e-06, | |
| "loss": 0.9282, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.8465238770975237e-06, | |
| "loss": 0.9574, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.8384180176991368e-06, | |
| "loss": 0.9312, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.8303017322641474e-06, | |
| "loss": 0.9416, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.8221752708403801e-06, | |
| "loss": 0.9267, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.8140388837891622e-06, | |
| "loss": 0.8971, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.8058928217776125e-06, | |
| "loss": 0.937, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.7977373357709162e-06, | |
| "loss": 0.9546, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.789572677024595e-06, | |
| "loss": 1.0132, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.7813990970767658e-06, | |
| "loss": 0.9453, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.7732168477403924e-06, | |
| "loss": 0.9584, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7650261810955248e-06, | |
| "loss": 0.9375, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.7568273494815353e-06, | |
| "loss": 0.9378, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7486206054893445e-06, | |
| "loss": 0.9765, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.7404062019536382e-06, | |
| "loss": 0.9465, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.7321843919450793e-06, | |
| "loss": 0.9333, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.7239554287625099e-06, | |
| "loss": 0.9359, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.715719565925148e-06, | |
| "loss": 0.9351, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.7074770571647776e-06, | |
| "loss": 1.0276, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.6992281564179327e-06, | |
| "loss": 0.95, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.69097311781807e-06, | |
| "loss": 0.9435, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.6827121956877436e-06, | |
| "loss": 0.9449, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.6744456445307693e-06, | |
| "loss": 0.9555, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.6661737190243813e-06, | |
| "loss": 0.9541, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.6578966740113882e-06, | |
| "loss": 0.9581, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.6496147644923206e-06, | |
| "loss": 0.9577, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.6413282456175774e-06, | |
| "loss": 0.9523, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.6330373726795605e-06, | |
| "loss": 0.9418, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.6247424011048153e-06, | |
| "loss": 0.9264, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.6164435864461566e-06, | |
| "loss": 0.9561, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.6081411843747983e-06, | |
| "loss": 0.9189, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.599835450672476e-06, | |
| "loss": 0.935, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5915266412235675e-06, | |
| "loss": 0.9408, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.5832150120072059e-06, | |
| "loss": 0.9134, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.5749008190893995e-06, | |
| "loss": 0.9113, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.5665843186151378e-06, | |
| "loss": 0.9446, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.5582657668005015e-06, | |
| "loss": 0.9548, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.5499454199247714e-06, | |
| "loss": 0.9186, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.541623534322528e-06, | |
| "loss": 0.9397, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.5333003663757585e-06, | |
| "loss": 0.9412, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.5249761725059577e-06, | |
| "loss": 0.9463, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.5166512091662264e-06, | |
| "loss": 0.9904, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.5083257328333697e-06, | |
| "loss": 0.9385, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.9535, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.4916742671666306e-06, | |
| "loss": 0.9471, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4833487908337741e-06, | |
| "loss": 0.9482, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.4750238274940423e-06, | |
| "loss": 0.943, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.4666996336242414e-06, | |
| "loss": 0.9973, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.4583764656774728e-06, | |
| "loss": 0.9453, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.4500545800752293e-06, | |
| "loss": 1.0162, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.4417342331994986e-06, | |
| "loss": 0.9118, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4334156813848625e-06, | |
| "loss": 0.9118, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.4250991809106006e-06, | |
| "loss": 0.9484, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.4167849879927944e-06, | |
| "loss": 0.9669, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.4084733587764328e-06, | |
| "loss": 1.0024, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4001645493275237e-06, | |
| "loss": 0.9233, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.3918588156252018e-06, | |
| "loss": 0.925, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.383556413553844e-06, | |
| "loss": 0.9361, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.3752575988951854e-06, | |
| "loss": 0.9187, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.36696262732044e-06, | |
| "loss": 1.0012, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.3586717543824231e-06, | |
| "loss": 0.928, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.3503852355076795e-06, | |
| "loss": 0.9427, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.3421033259886123e-06, | |
| "loss": 0.9211, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.3338262809756188e-06, | |
| "loss": 1.0299, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.3255543554692306e-06, | |
| "loss": 0.9101, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.3172878043122562e-06, | |
| "loss": 0.9423, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.3090268821819306e-06, | |
| "loss": 0.9319, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.300771843582068e-06, | |
| "loss": 0.9378, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.2925229428352225e-06, | |
| "loss": 0.9213, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.2842804340748524e-06, | |
| "loss": 0.9675, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.2760445712374906e-06, | |
| "loss": 0.9304, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.267815608054921e-06, | |
| "loss": 0.9335, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.2595937980463616e-06, | |
| "loss": 0.9215, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.2513793945106556e-06, | |
| "loss": 0.9341, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.2431726505184652e-06, | |
| "loss": 0.9299, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.234973818904476e-06, | |
| "loss": 0.9259, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.2267831522596081e-06, | |
| "loss": 0.935, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.2186009029232343e-06, | |
| "loss": 0.9496, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.2104273229754053e-06, | |
| "loss": 0.9341, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.2022626642290843e-06, | |
| "loss": 0.9318, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.194107178222388e-06, | |
| "loss": 0.9337, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.1859611162108379e-06, | |
| "loss": 0.9496, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.17782472915962e-06, | |
| "loss": 0.9327, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.1696982677358533e-06, | |
| "loss": 0.9576, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.1615819823008639e-06, | |
| "loss": 0.9423, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.153476122902477e-06, | |
| "loss": 0.991, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1453809392673101e-06, | |
| "loss": 0.9247, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.137296680793079e-06, | |
| "loss": 0.9365, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.1292235965409171e-06, | |
| "loss": 0.9076, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.1211619352276997e-06, | |
| "loss": 0.936, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.1131119452183814e-06, | |
| "loss": 0.9052, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.1050738745183472e-06, | |
| "loss": 0.9136, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.0970479707657699e-06, | |
| "loss": 0.9287, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.0890344812239785e-06, | |
| "loss": 0.9556, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.0810336527738461e-06, | |
| "loss": 0.9015, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.0730457319061797e-06, | |
| "loss": 0.9284, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.065070964714127e-06, | |
| "loss": 0.9271, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.0571095968855957e-06, | |
| "loss": 0.8869, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.049161873695684e-06, | |
| "loss": 0.9009, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.0412280399991226e-06, | |
| "loss": 0.9528, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.033308340222734e-06, | |
| "loss": 0.953, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.0254030183579003e-06, | |
| "loss": 0.931, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.0175123179530452e-06, | |
| "loss": 0.9086, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.0096364821061337e-06, | |
| "loss": 0.9338, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.0017757534571806e-06, | |
| "loss": 0.9267, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 9.939303741807755e-07, | |
| "loss": 0.8946, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.861005859786223e-07, | |
| "loss": 0.907, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 9.78286630072093e-07, | |
| "loss": 0.9366, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.70488747194795e-07, | |
| "loss": 0.9417, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.627071775851547e-07, | |
| "loss": 0.9247, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 9.549421609790195e-07, | |
| "loss": 0.9132, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 9.471939366022647e-07, | |
| "loss": 0.9296, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 9.394627431634321e-07, | |
| "loss": 0.9341, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.317488188463695e-07, | |
| "loss": 0.9266, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 9.240524013028948e-07, | |
| "loss": 0.9269, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 9.163737276454758e-07, | |
| "loss": 0.9158, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 9.087130344399223e-07, | |
| "loss": 0.8931, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 9.010705576981002e-07, | |
| "loss": 0.9486, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.934465328706593e-07, | |
| "loss": 0.9348, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.858411948397813e-07, | |
| "loss": 0.9305, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.782547779119386e-07, | |
| "loss": 0.9138, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.706875158106834e-07, | |
| "loss": 0.9354, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.6313964166944e-07, | |
| "loss": 0.9169, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.556113880243266e-07, | |
| "loss": 0.9358, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.481029868069898e-07, | |
| "loss": 0.9501, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.406146693374587e-07, | |
| "loss": 0.9193, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.331466663170208e-07, | |
| "loss": 1.0219, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 8.256992078211112e-07, | |
| "loss": 0.9451, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 8.182725232922269e-07, | |
| "loss": 0.9171, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.10866841532856e-07, | |
| "loss": 0.923, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.034823906984308e-07, | |
| "loss": 0.9226, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 7.961193982902977e-07, | |
| "loss": 0.9091, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 7.88778091148709e-07, | |
| "loss": 0.9069, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 7.814586954458334e-07, | |
| "loss": 0.9241, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.741614366787881e-07, | |
| "loss": 0.9249, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 7.668865396626924e-07, | |
| "loss": 0.9151, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.59634228523741e-07, | |
| "loss": 0.9422, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 7.524047266922997e-07, | |
| "loss": 0.9168, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.451982568960207e-07, | |
| "loss": 0.9353, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.380150411529826e-07, | |
| "loss": 0.9208, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.308553007648485e-07, | |
| "loss": 0.9216, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.237192563100496e-07, | |
| "loss": 0.973, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.166071276369886e-07, | |
| "loss": 0.8989, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.095191338572666e-07, | |
| "loss": 0.909, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 7.024554933389344e-07, | |
| "loss": 0.9199, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 6.95416423699763e-07, | |
| "loss": 0.9297, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 6.884021418005384e-07, | |
| "loss": 0.9304, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.814128637383837e-07, | |
| "loss": 0.9458, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.74448804840099e-07, | |
| "loss": 0.9615, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 6.675101796555279e-07, | |
| "loss": 0.9203, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.605972019509501e-07, | |
| "loss": 0.9297, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 6.537100847024914e-07, | |
| "loss": 0.9314, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 6.468490400895653e-07, | |
| "loss": 0.9189, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 6.400142794883356e-07, | |
| "loss": 0.9125, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.332060134652033e-07, | |
| "loss": 0.9308, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.264244517703215e-07, | |
| "loss": 0.9439, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 6.196698033311305e-07, | |
| "loss": 0.9444, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 6.12942276245924e-07, | |
| "loss": 0.9637, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.0, | |
| "learning_rate": 6.062420777774359e-07, | |
| "loss": 0.8928, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.99569414346456e-07, | |
| "loss": 0.9321, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.929244915254703e-07, | |
| "loss": 0.8973, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.86307514032327e-07, | |
| "loss": 0.9113, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.797186857239313e-07, | |
| "loss": 0.9453, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.731582095899636e-07, | |
| "loss": 0.9417, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 5.666262877466246e-07, | |
| "loss": 0.9198, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.601231214304107e-07, | |
| "loss": 0.9268, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 2.0, | |
| "learning_rate": 5.536489109919141e-07, | |
| "loss": 0.9272, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.472038558896483e-07, | |
| "loss": 0.9091, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.40788154683906e-07, | |
| "loss": 0.9193, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.344020050306396e-07, | |
| "loss": 0.9212, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.280456036753723e-07, | |
| "loss": 1.0223, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.217191464471373e-07, | |
| "loss": 0.9886, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.15422828252444e-07, | |
| "loss": 0.9164, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.091568430692738e-07, | |
| "loss": 0.9474, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.029213839411043e-07, | |
| "loss": 0.9271, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.967166429709606e-07, | |
| "loss": 0.9369, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.905428113154986e-07, | |
| "loss": 0.9415, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 4.844000791791147e-07, | |
| "loss": 0.9494, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.782886358080865e-07, | |
| "loss": 0.8931, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.7220866948474156e-07, | |
| "loss": 0.922, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.6616036752165916e-07, | |
| "loss": 0.9282, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.6014391625589697e-07, | |
| "loss": 0.9364, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.541595010432501e-07, | |
| "loss": 0.9398, | |
| "step": 432 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 576, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 72, | |
| "total_flos": 6.374281953222328e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |