| { | |
| "best_metric": 1.542069435119629, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.47095761381475665, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0031397174254317113, | |
| "grad_norm": 6.254443645477295, | |
| "learning_rate": 1.0140000000000001e-05, | |
| "loss": 1.6655, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0031397174254317113, | |
| "eval_loss": 3.4825549125671387, | |
| "eval_runtime": 3.7073, | |
| "eval_samples_per_second": 36.145, | |
| "eval_steps_per_second": 9.171, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006279434850863423, | |
| "grad_norm": 7.797785758972168, | |
| "learning_rate": 2.0280000000000002e-05, | |
| "loss": 2.381, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009419152276295133, | |
| "grad_norm": 8.442407608032227, | |
| "learning_rate": 3.0419999999999997e-05, | |
| "loss": 2.405, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.012558869701726845, | |
| "grad_norm": 8.922061920166016, | |
| "learning_rate": 4.0560000000000005e-05, | |
| "loss": 2.7349, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.015698587127158554, | |
| "grad_norm": 8.771944999694824, | |
| "learning_rate": 5.07e-05, | |
| "loss": 2.5019, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.018838304552590265, | |
| "grad_norm": 7.558984756469727, | |
| "learning_rate": 6.0839999999999993e-05, | |
| "loss": 2.3498, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02197802197802198, | |
| "grad_norm": 7.375399589538574, | |
| "learning_rate": 7.097999999999999e-05, | |
| "loss": 1.9931, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02511773940345369, | |
| "grad_norm": 8.363454818725586, | |
| "learning_rate": 8.112000000000001e-05, | |
| "loss": 2.4119, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0282574568288854, | |
| "grad_norm": 8.049056053161621, | |
| "learning_rate": 9.126e-05, | |
| "loss": 2.1919, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03139717425431711, | |
| "grad_norm": 7.283090114593506, | |
| "learning_rate": 0.0001014, | |
| "loss": 1.694, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03453689167974882, | |
| "grad_norm": 7.796234130859375, | |
| "learning_rate": 0.00010086631578947368, | |
| "loss": 1.8361, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03767660910518053, | |
| "grad_norm": 8.134119033813477, | |
| "learning_rate": 0.00010033263157894736, | |
| "loss": 2.4029, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 10.318169593811035, | |
| "learning_rate": 9.979894736842105e-05, | |
| "loss": 2.4189, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04395604395604396, | |
| "grad_norm": 7.024040222167969, | |
| "learning_rate": 9.926526315789475e-05, | |
| "loss": 1.3956, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04709576138147567, | |
| "grad_norm": 7.71487283706665, | |
| "learning_rate": 9.873157894736843e-05, | |
| "loss": 1.8253, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05023547880690738, | |
| "grad_norm": 8.051153182983398, | |
| "learning_rate": 9.81978947368421e-05, | |
| "loss": 1.9171, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05337519623233909, | |
| "grad_norm": 9.367609977722168, | |
| "learning_rate": 9.766421052631579e-05, | |
| "loss": 1.918, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0565149136577708, | |
| "grad_norm": 7.217973709106445, | |
| "learning_rate": 9.713052631578947e-05, | |
| "loss": 1.5383, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.059654631083202514, | |
| "grad_norm": 8.168498992919922, | |
| "learning_rate": 9.659684210526315e-05, | |
| "loss": 1.9243, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 9.34277629852295, | |
| "learning_rate": 9.606315789473684e-05, | |
| "loss": 1.9905, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06593406593406594, | |
| "grad_norm": 9.115056037902832, | |
| "learning_rate": 9.552947368421053e-05, | |
| "loss": 1.9055, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06907378335949764, | |
| "grad_norm": 10.160658836364746, | |
| "learning_rate": 9.499578947368422e-05, | |
| "loss": 1.8786, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07221350078492936, | |
| "grad_norm": 8.394580841064453, | |
| "learning_rate": 9.44621052631579e-05, | |
| "loss": 1.7636, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07535321821036106, | |
| "grad_norm": 7.4472832679748535, | |
| "learning_rate": 9.392842105263158e-05, | |
| "loss": 1.476, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.07849293563579278, | |
| "grad_norm": 9.046372413635254, | |
| "learning_rate": 9.339473684210526e-05, | |
| "loss": 2.6342, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 12.512232780456543, | |
| "learning_rate": 9.286105263157894e-05, | |
| "loss": 2.0217, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0847723704866562, | |
| "grad_norm": 7.891317844390869, | |
| "learning_rate": 9.232736842105263e-05, | |
| "loss": 1.8033, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.08791208791208792, | |
| "grad_norm": 8.680127143859863, | |
| "learning_rate": 9.179368421052632e-05, | |
| "loss": 1.8667, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09105180533751962, | |
| "grad_norm": 8.955038070678711, | |
| "learning_rate": 9.126e-05, | |
| "loss": 1.8639, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.09419152276295134, | |
| "grad_norm": 8.710933685302734, | |
| "learning_rate": 9.072631578947368e-05, | |
| "loss": 1.9586, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09733124018838304, | |
| "grad_norm": 7.248199939727783, | |
| "learning_rate": 9.019263157894736e-05, | |
| "loss": 1.5127, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10047095761381476, | |
| "grad_norm": 7.970192909240723, | |
| "learning_rate": 8.965894736842104e-05, | |
| "loss": 1.5677, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10361067503924647, | |
| "grad_norm": 9.28384017944336, | |
| "learning_rate": 8.912526315789472e-05, | |
| "loss": 1.5312, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.10675039246467818, | |
| "grad_norm": 9.90267276763916, | |
| "learning_rate": 8.859157894736842e-05, | |
| "loss": 2.2637, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.10989010989010989, | |
| "grad_norm": 9.363068580627441, | |
| "learning_rate": 8.805789473684211e-05, | |
| "loss": 2.0247, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1130298273155416, | |
| "grad_norm": 9.736919403076172, | |
| "learning_rate": 8.752421052631579e-05, | |
| "loss": 1.6372, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.11616954474097331, | |
| "grad_norm": 9.847123146057129, | |
| "learning_rate": 8.699052631578947e-05, | |
| "loss": 1.9869, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.11930926216640503, | |
| "grad_norm": 10.517101287841797, | |
| "learning_rate": 8.645684210526315e-05, | |
| "loss": 2.3098, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 12.2364501953125, | |
| "learning_rate": 8.592315789473683e-05, | |
| "loss": 2.0425, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 8.809972763061523, | |
| "learning_rate": 8.538947368421051e-05, | |
| "loss": 2.2248, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12872841444270017, | |
| "grad_norm": 9.04055118560791, | |
| "learning_rate": 8.485578947368421e-05, | |
| "loss": 2.1993, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.13186813186813187, | |
| "grad_norm": 12.418622016906738, | |
| "learning_rate": 8.43221052631579e-05, | |
| "loss": 1.8935, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.13500784929356358, | |
| "grad_norm": 12.07343578338623, | |
| "learning_rate": 8.378842105263158e-05, | |
| "loss": 2.5241, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.13814756671899528, | |
| "grad_norm": 12.244543075561523, | |
| "learning_rate": 8.325473684210526e-05, | |
| "loss": 1.9638, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.141287284144427, | |
| "grad_norm": 11.888532638549805, | |
| "learning_rate": 8.272105263157894e-05, | |
| "loss": 2.2361, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.14442700156985872, | |
| "grad_norm": 9.17966365814209, | |
| "learning_rate": 8.218736842105262e-05, | |
| "loss": 1.7368, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.14756671899529042, | |
| "grad_norm": 12.1638822555542, | |
| "learning_rate": 8.165368421052632e-05, | |
| "loss": 2.0467, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.15070643642072212, | |
| "grad_norm": 11.597208023071289, | |
| "learning_rate": 8.112000000000001e-05, | |
| "loss": 1.7656, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 11.913068771362305, | |
| "learning_rate": 8.058631578947369e-05, | |
| "loss": 2.2715, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.15698587127158556, | |
| "grad_norm": 17.56863784790039, | |
| "learning_rate": 8.005263157894737e-05, | |
| "loss": 2.0193, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15698587127158556, | |
| "eval_loss": 1.7894538640975952, | |
| "eval_runtime": 3.7091, | |
| "eval_samples_per_second": 36.128, | |
| "eval_steps_per_second": 9.167, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16012558869701726, | |
| "grad_norm": 5.96769905090332, | |
| "learning_rate": 7.951894736842105e-05, | |
| "loss": 1.458, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 6.817239284515381, | |
| "learning_rate": 7.898526315789473e-05, | |
| "loss": 1.401, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1664050235478807, | |
| "grad_norm": 7.3476033210754395, | |
| "learning_rate": 7.845157894736841e-05, | |
| "loss": 1.8913, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1695447409733124, | |
| "grad_norm": 6.681586742401123, | |
| "learning_rate": 7.79178947368421e-05, | |
| "loss": 1.9426, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1726844583987441, | |
| "grad_norm": 6.83116340637207, | |
| "learning_rate": 7.73842105263158e-05, | |
| "loss": 2.0575, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.17582417582417584, | |
| "grad_norm": 8.655893325805664, | |
| "learning_rate": 7.685052631578948e-05, | |
| "loss": 1.8735, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.17896389324960754, | |
| "grad_norm": 7.9880194664001465, | |
| "learning_rate": 7.631684210526316e-05, | |
| "loss": 1.6114, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.18210361067503925, | |
| "grad_norm": 5.3605055809021, | |
| "learning_rate": 7.578315789473684e-05, | |
| "loss": 1.5865, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.18524332810047095, | |
| "grad_norm": 5.455999374389648, | |
| "learning_rate": 7.524947368421052e-05, | |
| "loss": 1.9276, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 6.868194103240967, | |
| "learning_rate": 7.47157894736842e-05, | |
| "loss": 2.0298, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.19152276295133439, | |
| "grad_norm": 5.856205940246582, | |
| "learning_rate": 7.418210526315789e-05, | |
| "loss": 1.7768, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1946624803767661, | |
| "grad_norm": 5.616762161254883, | |
| "learning_rate": 7.364842105263159e-05, | |
| "loss": 1.8175, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1978021978021978, | |
| "grad_norm": 6.230925559997559, | |
| "learning_rate": 7.311473684210527e-05, | |
| "loss": 1.3076, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.20094191522762953, | |
| "grad_norm": 6.4075798988342285, | |
| "learning_rate": 7.258105263157895e-05, | |
| "loss": 1.6211, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 5.425053119659424, | |
| "learning_rate": 7.204736842105263e-05, | |
| "loss": 1.5966, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.20722135007849293, | |
| "grad_norm": 6.285729885101318, | |
| "learning_rate": 7.151368421052631e-05, | |
| "loss": 1.9486, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.21036106750392464, | |
| "grad_norm": 8.634322166442871, | |
| "learning_rate": 7.097999999999999e-05, | |
| "loss": 1.6922, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.21350078492935637, | |
| "grad_norm": 7.055955410003662, | |
| "learning_rate": 7.044631578947368e-05, | |
| "loss": 2.0679, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.21664050235478807, | |
| "grad_norm": 6.999521732330322, | |
| "learning_rate": 6.991263157894738e-05, | |
| "loss": 2.0098, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.21978021978021978, | |
| "grad_norm": 6.568216800689697, | |
| "learning_rate": 6.937894736842106e-05, | |
| "loss": 1.6266, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.22291993720565148, | |
| "grad_norm": 6.6800689697265625, | |
| "learning_rate": 6.884526315789474e-05, | |
| "loss": 1.672, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2260596546310832, | |
| "grad_norm": 6.454991340637207, | |
| "learning_rate": 6.831157894736842e-05, | |
| "loss": 1.5975, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.22919937205651492, | |
| "grad_norm": 5.533199310302734, | |
| "learning_rate": 6.77778947368421e-05, | |
| "loss": 1.2333, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.23233908948194662, | |
| "grad_norm": 5.433882713317871, | |
| "learning_rate": 6.724421052631579e-05, | |
| "loss": 1.8005, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.23547880690737832, | |
| "grad_norm": 7.069916248321533, | |
| "learning_rate": 6.671052631578948e-05, | |
| "loss": 1.8686, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.23861852433281006, | |
| "grad_norm": 6.151008605957031, | |
| "learning_rate": 6.617684210526316e-05, | |
| "loss": 1.4558, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.24175824175824176, | |
| "grad_norm": 5.1850104331970215, | |
| "learning_rate": 6.564315789473684e-05, | |
| "loss": 1.3295, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 5.7544331550598145, | |
| "learning_rate": 6.510947368421052e-05, | |
| "loss": 1.3047, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.24803767660910517, | |
| "grad_norm": 6.109000205993652, | |
| "learning_rate": 6.45757894736842e-05, | |
| "loss": 1.4995, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 6.1000566482543945, | |
| "learning_rate": 6.404210526315789e-05, | |
| "loss": 1.1538, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2543171114599686, | |
| "grad_norm": 6.75916862487793, | |
| "learning_rate": 6.350842105263158e-05, | |
| "loss": 1.6112, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.25745682888540034, | |
| "grad_norm": 5.901347637176514, | |
| "learning_rate": 6.297473684210527e-05, | |
| "loss": 1.2695, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.260596546310832, | |
| "grad_norm": 6.797555923461914, | |
| "learning_rate": 6.244105263157895e-05, | |
| "loss": 1.7123, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.26373626373626374, | |
| "grad_norm": 8.325261116027832, | |
| "learning_rate": 6.190736842105263e-05, | |
| "loss": 1.5652, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2668759811616955, | |
| "grad_norm": 7.468142509460449, | |
| "learning_rate": 6.137368421052631e-05, | |
| "loss": 1.85, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.27001569858712715, | |
| "grad_norm": 6.591063499450684, | |
| "learning_rate": 6.0839999999999993e-05, | |
| "loss": 1.2159, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2731554160125589, | |
| "grad_norm": 7.766685962677002, | |
| "learning_rate": 6.030631578947368e-05, | |
| "loss": 1.9582, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.27629513343799056, | |
| "grad_norm": 7.326028347015381, | |
| "learning_rate": 5.977263157894736e-05, | |
| "loss": 1.6978, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2794348508634223, | |
| "grad_norm": 7.96261739730835, | |
| "learning_rate": 5.9238947368421054e-05, | |
| "loss": 1.7344, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.282574568288854, | |
| "grad_norm": 6.848148822784424, | |
| "learning_rate": 5.870526315789474e-05, | |
| "loss": 1.5061, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 6.8328752517700195, | |
| "learning_rate": 5.817157894736842e-05, | |
| "loss": 1.7432, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.28885400313971743, | |
| "grad_norm": 11.828781127929688, | |
| "learning_rate": 5.76378947368421e-05, | |
| "loss": 2.2161, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.29199372056514916, | |
| "grad_norm": 8.98495101928711, | |
| "learning_rate": 5.710421052631579e-05, | |
| "loss": 1.9033, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.29513343799058084, | |
| "grad_norm": 16.75680160522461, | |
| "learning_rate": 5.657052631578947e-05, | |
| "loss": 2.0846, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.29827315541601257, | |
| "grad_norm": 10.899314880371094, | |
| "learning_rate": 5.603684210526316e-05, | |
| "loss": 1.7966, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.30141287284144425, | |
| "grad_norm": 8.247507095336914, | |
| "learning_rate": 5.550315789473684e-05, | |
| "loss": 1.4718, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.304552590266876, | |
| "grad_norm": 10.83910846710205, | |
| "learning_rate": 5.496947368421053e-05, | |
| "loss": 1.6901, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 9.814567565917969, | |
| "learning_rate": 5.443578947368421e-05, | |
| "loss": 2.1697, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3108320251177394, | |
| "grad_norm": 11.087676048278809, | |
| "learning_rate": 5.390210526315789e-05, | |
| "loss": 1.9234, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 16.154081344604492, | |
| "learning_rate": 5.336842105263158e-05, | |
| "loss": 2.2297, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "eval_loss": 1.6598618030548096, | |
| "eval_runtime": 3.7012, | |
| "eval_samples_per_second": 36.204, | |
| "eval_steps_per_second": 9.186, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.31711145996860285, | |
| "grad_norm": 3.5819578170776367, | |
| "learning_rate": 5.283473684210526e-05, | |
| "loss": 1.0256, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3202511773940345, | |
| "grad_norm": 6.239628314971924, | |
| "learning_rate": 5.230105263157895e-05, | |
| "loss": 2.0302, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.32339089481946626, | |
| "grad_norm": 6.583393573760986, | |
| "learning_rate": 5.176736842105263e-05, | |
| "loss": 1.8993, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 6.852742671966553, | |
| "learning_rate": 5.123368421052632e-05, | |
| "loss": 1.5832, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.32967032967032966, | |
| "grad_norm": 6.229090690612793, | |
| "learning_rate": 5.07e-05, | |
| "loss": 1.7996, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3328100470957614, | |
| "grad_norm": 4.9963908195495605, | |
| "learning_rate": 5.016631578947368e-05, | |
| "loss": 1.3517, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3359497645211931, | |
| "grad_norm": 6.220682144165039, | |
| "learning_rate": 4.963263157894737e-05, | |
| "loss": 2.3754, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3390894819466248, | |
| "grad_norm": 5.58546781539917, | |
| "learning_rate": 4.909894736842105e-05, | |
| "loss": 1.9693, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.34222919937205654, | |
| "grad_norm": 5.197764873504639, | |
| "learning_rate": 4.8565263157894734e-05, | |
| "loss": 1.1676, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3453689167974882, | |
| "grad_norm": 6.040829181671143, | |
| "learning_rate": 4.803157894736842e-05, | |
| "loss": 1.9262, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.34850863422291994, | |
| "grad_norm": 6.036067485809326, | |
| "learning_rate": 4.749789473684211e-05, | |
| "loss": 1.875, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3516483516483517, | |
| "grad_norm": 5.573509693145752, | |
| "learning_rate": 4.696421052631579e-05, | |
| "loss": 1.5932, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.35478806907378335, | |
| "grad_norm": 5.3029632568359375, | |
| "learning_rate": 4.643052631578947e-05, | |
| "loss": 1.3575, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3579277864992151, | |
| "grad_norm": 5.437764644622803, | |
| "learning_rate": 4.589684210526316e-05, | |
| "loss": 1.6235, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.36106750392464676, | |
| "grad_norm": 4.923410415649414, | |
| "learning_rate": 4.536315789473684e-05, | |
| "loss": 1.3825, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3642072213500785, | |
| "grad_norm": 6.01725435256958, | |
| "learning_rate": 4.482947368421052e-05, | |
| "loss": 1.4735, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 6.325821876525879, | |
| "learning_rate": 4.429578947368421e-05, | |
| "loss": 1.6234, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3704866562009419, | |
| "grad_norm": 5.018939018249512, | |
| "learning_rate": 4.3762105263157896e-05, | |
| "loss": 1.7545, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.37362637362637363, | |
| "grad_norm": 5.737129211425781, | |
| "learning_rate": 4.3228421052631576e-05, | |
| "loss": 1.7748, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 5.9904866218566895, | |
| "learning_rate": 4.269473684210526e-05, | |
| "loss": 1.4207, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.37990580847723704, | |
| "grad_norm": 7.162813186645508, | |
| "learning_rate": 4.216105263157895e-05, | |
| "loss": 1.5775, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.38304552590266877, | |
| "grad_norm": 4.571135520935059, | |
| "learning_rate": 4.162736842105263e-05, | |
| "loss": 1.6661, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.38618524332810045, | |
| "grad_norm": 5.065517902374268, | |
| "learning_rate": 4.109368421052631e-05, | |
| "loss": 1.5862, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3893249607535322, | |
| "grad_norm": 7.399703025817871, | |
| "learning_rate": 4.0560000000000005e-05, | |
| "loss": 2.1326, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3924646781789639, | |
| "grad_norm": 5.505315780639648, | |
| "learning_rate": 4.0026315789473685e-05, | |
| "loss": 1.265, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3956043956043956, | |
| "grad_norm": 6.12184476852417, | |
| "learning_rate": 3.9492631578947365e-05, | |
| "loss": 1.5402, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3987441130298273, | |
| "grad_norm": 6.294252395629883, | |
| "learning_rate": 3.895894736842105e-05, | |
| "loss": 1.5752, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.40188383045525905, | |
| "grad_norm": 6.789506435394287, | |
| "learning_rate": 3.842526315789474e-05, | |
| "loss": 1.9103, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4050235478806907, | |
| "grad_norm": 6.3864006996154785, | |
| "learning_rate": 3.789157894736842e-05, | |
| "loss": 1.7966, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 7.398538589477539, | |
| "learning_rate": 3.73578947368421e-05, | |
| "loss": 1.9872, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.41130298273155413, | |
| "grad_norm": 7.464357376098633, | |
| "learning_rate": 3.682421052631579e-05, | |
| "loss": 1.8161, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.41444270015698587, | |
| "grad_norm": 7.386998176574707, | |
| "learning_rate": 3.6290526315789474e-05, | |
| "loss": 1.5991, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4175824175824176, | |
| "grad_norm": 7.29685115814209, | |
| "learning_rate": 3.5756842105263154e-05, | |
| "loss": 1.7373, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4207221350078493, | |
| "grad_norm": 6.442423343658447, | |
| "learning_rate": 3.522315789473684e-05, | |
| "loss": 1.5062, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.423861852433281, | |
| "grad_norm": 7.622251987457275, | |
| "learning_rate": 3.468947368421053e-05, | |
| "loss": 1.6362, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.42700156985871274, | |
| "grad_norm": 5.391417503356934, | |
| "learning_rate": 3.415578947368421e-05, | |
| "loss": 1.3335, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4301412872841444, | |
| "grad_norm": 7.495037078857422, | |
| "learning_rate": 3.3622105263157895e-05, | |
| "loss": 2.1399, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.43328100470957615, | |
| "grad_norm": 8.007742881774902, | |
| "learning_rate": 3.308842105263158e-05, | |
| "loss": 1.8728, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4364207221350079, | |
| "grad_norm": 5.115504741668701, | |
| "learning_rate": 3.255473684210526e-05, | |
| "loss": 1.3832, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 7.03964376449585, | |
| "learning_rate": 3.202105263157894e-05, | |
| "loss": 1.8599, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4427001569858713, | |
| "grad_norm": 7.27022647857666, | |
| "learning_rate": 3.1487368421052636e-05, | |
| "loss": 1.5044, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.44583987441130296, | |
| "grad_norm": 8.688240051269531, | |
| "learning_rate": 3.0953684210526317e-05, | |
| "loss": 1.9889, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 7.870269298553467, | |
| "learning_rate": 3.0419999999999997e-05, | |
| "loss": 1.5771, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4521193092621664, | |
| "grad_norm": 6.881030082702637, | |
| "learning_rate": 2.988631578947368e-05, | |
| "loss": 1.2406, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4552590266875981, | |
| "grad_norm": 7.38179349899292, | |
| "learning_rate": 2.935263157894737e-05, | |
| "loss": 1.5986, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.45839874411302983, | |
| "grad_norm": 7.3739800453186035, | |
| "learning_rate": 2.881894736842105e-05, | |
| "loss": 1.5282, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 8.048142433166504, | |
| "learning_rate": 2.8285263157894735e-05, | |
| "loss": 1.7917, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.46467817896389324, | |
| "grad_norm": 8.414251327514648, | |
| "learning_rate": 2.775157894736842e-05, | |
| "loss": 2.0077, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.46781789638932497, | |
| "grad_norm": 10.131597518920898, | |
| "learning_rate": 2.7217894736842105e-05, | |
| "loss": 2.0688, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.47095761381475665, | |
| "grad_norm": 13.612863540649414, | |
| "learning_rate": 2.668421052631579e-05, | |
| "loss": 1.7537, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.47095761381475665, | |
| "eval_loss": 1.542069435119629, | |
| "eval_runtime": 3.6868, | |
| "eval_samples_per_second": 36.346, | |
| "eval_steps_per_second": 9.222, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2898180911923200.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |