dialect-debiasing-gpt2-medium-translated-pnlogmse-e1-r10_eval-n10.0
/
checkpoint-3125
/trainer_state.json
| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 10.000000953674316, | |
| "learning_rate": 6.369426751592357e-06, | |
| "loss": 22.4391, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 1.2738853503184714e-05, | |
| "loss": 18.5398, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 10.0, | |
| "learning_rate": 1.910828025477707e-05, | |
| "loss": 18.8696, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 2.5477707006369428e-05, | |
| "loss": 17.9832, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.184713375796178e-05, | |
| "loss": 15.0181, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 3.821656050955414e-05, | |
| "loss": 14.2746, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 10.0, | |
| "learning_rate": 4.45859872611465e-05, | |
| "loss": 11.9335, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 10.0, | |
| "learning_rate": 5.0955414012738855e-05, | |
| "loss": 9.6611, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 10.0, | |
| "learning_rate": 5.732484076433121e-05, | |
| "loss": 8.2423, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 6.369426751592356e-05, | |
| "loss": 7.0975, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 10.0, | |
| "learning_rate": 7.006369426751592e-05, | |
| "loss": 5.2427, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 7.643312101910829e-05, | |
| "loss": 5.0576, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 8.280254777070065e-05, | |
| "loss": 4.3215, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 9.999999046325684, | |
| "learning_rate": 8.9171974522293e-05, | |
| "loss": 3.7479, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 9.422039985656738, | |
| "learning_rate": 9.554140127388536e-05, | |
| "loss": 3.5213, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 9.999998092651367, | |
| "learning_rate": 9.989892183288411e-05, | |
| "loss": 3.2425, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 9.351380348205566, | |
| "learning_rate": 9.956199460916442e-05, | |
| "loss": 2.8891, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 8.307005882263184, | |
| "learning_rate": 9.922506738544474e-05, | |
| "loss": 2.783, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 8.963129997253418, | |
| "learning_rate": 9.888814016172507e-05, | |
| "loss": 2.7527, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 8.570749282836914, | |
| "learning_rate": 9.85512129380054e-05, | |
| "loss": 2.4355, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 8.063360214233398, | |
| "learning_rate": 9.821428571428572e-05, | |
| "loss": 2.4091, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 8.824004173278809, | |
| "learning_rate": 9.787735849056603e-05, | |
| "loss": 2.1608, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 7.373661994934082, | |
| "learning_rate": 9.754043126684636e-05, | |
| "loss": 2.2534, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 8.307866096496582, | |
| "learning_rate": 9.720350404312669e-05, | |
| "loss": 2.1537, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.856404781341553, | |
| "learning_rate": 9.686657681940702e-05, | |
| "loss": 1.9916, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 8.64548397064209, | |
| "learning_rate": 9.652964959568734e-05, | |
| "loss": 1.9904, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 7.271899223327637, | |
| "learning_rate": 9.619272237196765e-05, | |
| "loss": 2.0503, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 6.368246555328369, | |
| "learning_rate": 9.585579514824798e-05, | |
| "loss": 1.9192, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 6.038721561431885, | |
| "learning_rate": 9.551886792452831e-05, | |
| "loss": 1.6597, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 6.260958194732666, | |
| "learning_rate": 9.518194070080863e-05, | |
| "loss": 1.9203, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 7.62371301651001, | |
| "learning_rate": 9.484501347708896e-05, | |
| "loss": 1.8547, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 7.491437911987305, | |
| "learning_rate": 9.450808625336927e-05, | |
| "loss": 1.7249, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 7.197510719299316, | |
| "learning_rate": 9.41711590296496e-05, | |
| "loss": 1.5601, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 6.928409099578857, | |
| "learning_rate": 9.383423180592993e-05, | |
| "loss": 1.6026, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 6.700569152832031, | |
| "learning_rate": 9.349730458221025e-05, | |
| "loss": 1.5981, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 6.20810604095459, | |
| "learning_rate": 9.316037735849057e-05, | |
| "loss": 1.4863, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 6.265394687652588, | |
| "learning_rate": 9.282345013477089e-05, | |
| "loss": 1.5054, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 7.249783039093018, | |
| "learning_rate": 9.248652291105122e-05, | |
| "loss": 1.4892, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 6.698989391326904, | |
| "learning_rate": 9.214959568733154e-05, | |
| "loss": 1.3346, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 6.408270359039307, | |
| "learning_rate": 9.181266846361186e-05, | |
| "loss": 1.269, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 6.221065521240234, | |
| "learning_rate": 9.14757412398922e-05, | |
| "loss": 1.2479, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 6.697783470153809, | |
| "learning_rate": 9.113881401617251e-05, | |
| "loss": 1.1823, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 6.298666000366211, | |
| "learning_rate": 9.080188679245284e-05, | |
| "loss": 1.2215, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 5.8949875831604, | |
| "learning_rate": 9.046495956873315e-05, | |
| "loss": 1.2094, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 4.865890026092529, | |
| "learning_rate": 9.012803234501348e-05, | |
| "loss": 1.1668, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 4.905447006225586, | |
| "learning_rate": 8.979110512129381e-05, | |
| "loss": 1.0941, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 5.108652591705322, | |
| "learning_rate": 8.945417789757413e-05, | |
| "loss": 1.081, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 4.898155212402344, | |
| "learning_rate": 8.911725067385444e-05, | |
| "loss": 1.0708, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 4.2186455726623535, | |
| "learning_rate": 8.878032345013477e-05, | |
| "loss": 0.9574, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.5638957023620605, | |
| "learning_rate": 8.84433962264151e-05, | |
| "loss": 1.1016, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 5.078132152557373, | |
| "learning_rate": 8.810646900269543e-05, | |
| "loss": 0.9938, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 6.263752460479736, | |
| "learning_rate": 8.776954177897575e-05, | |
| "loss": 1.0042, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.1696, | |
| "grad_norm": 4.613877773284912, | |
| "learning_rate": 8.743261455525606e-05, | |
| "loss": 0.8717, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 5.016404151916504, | |
| "learning_rate": 8.709568733153639e-05, | |
| "loss": 0.9252, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 4.927691459655762, | |
| "learning_rate": 8.675876010781672e-05, | |
| "loss": 0.9297, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 4.506190299987793, | |
| "learning_rate": 8.642183288409704e-05, | |
| "loss": 0.8883, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 5.3836350440979, | |
| "learning_rate": 8.608490566037735e-05, | |
| "loss": 0.8302, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 4.673761367797852, | |
| "learning_rate": 8.574797843665768e-05, | |
| "loss": 0.8401, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.1888, | |
| "grad_norm": 3.7030556201934814, | |
| "learning_rate": 8.541105121293801e-05, | |
| "loss": 0.888, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 4.952301025390625, | |
| "learning_rate": 8.507412398921834e-05, | |
| "loss": 0.8483, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1952, | |
| "grad_norm": 4.4545769691467285, | |
| "learning_rate": 8.473719676549866e-05, | |
| "loss": 0.7925, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 4.0674896240234375, | |
| "learning_rate": 8.440026954177897e-05, | |
| "loss": 0.7347, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 6.8568034172058105, | |
| "learning_rate": 8.40633423180593e-05, | |
| "loss": 0.7413, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 6.715765476226807, | |
| "learning_rate": 8.372641509433963e-05, | |
| "loss": 0.7714, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 4.419723033905029, | |
| "learning_rate": 8.338948787061996e-05, | |
| "loss": 0.7337, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 4.291261672973633, | |
| "learning_rate": 8.305256064690027e-05, | |
| "loss": 0.6854, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.2144, | |
| "grad_norm": 5.974247932434082, | |
| "learning_rate": 8.271563342318059e-05, | |
| "loss": 0.7224, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 4.334232807159424, | |
| "learning_rate": 8.237870619946092e-05, | |
| "loss": 0.6737, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 4.699735164642334, | |
| "learning_rate": 8.204177897574125e-05, | |
| "loss": 0.6505, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 4.12949275970459, | |
| "learning_rate": 8.170485175202158e-05, | |
| "loss": 0.6696, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2272, | |
| "grad_norm": 3.9896364212036133, | |
| "learning_rate": 8.136792452830189e-05, | |
| "loss": 0.6587, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 4.438923358917236, | |
| "learning_rate": 8.103099730458221e-05, | |
| "loss": 0.6189, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.2336, | |
| "grad_norm": 3.4031989574432373, | |
| "learning_rate": 8.069407008086254e-05, | |
| "loss": 0.6249, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 5.202292442321777, | |
| "learning_rate": 8.035714285714287e-05, | |
| "loss": 0.6462, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 5.27857780456543, | |
| "learning_rate": 8.002021563342318e-05, | |
| "loss": 0.5842, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 4.678514003753662, | |
| "learning_rate": 7.968328840970351e-05, | |
| "loss": 0.5659, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.2464, | |
| "grad_norm": 4.691657543182373, | |
| "learning_rate": 7.934636118598383e-05, | |
| "loss": 0.5833, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 7.131917476654053, | |
| "learning_rate": 7.900943396226416e-05, | |
| "loss": 0.5958, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2528, | |
| "grad_norm": 5.070577144622803, | |
| "learning_rate": 7.867250673854449e-05, | |
| "loss": 0.6307, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 4.827574253082275, | |
| "learning_rate": 7.83355795148248e-05, | |
| "loss": 0.6064, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 3.8818304538726807, | |
| "learning_rate": 7.799865229110512e-05, | |
| "loss": 0.5218, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 6.581722736358643, | |
| "learning_rate": 7.766172506738545e-05, | |
| "loss": 0.5342, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.2656, | |
| "grad_norm": 9.561187744140625, | |
| "learning_rate": 7.732479784366577e-05, | |
| "loss": 0.5167, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 6.698048114776611, | |
| "learning_rate": 7.69878706199461e-05, | |
| "loss": 0.5674, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 5.142230987548828, | |
| "learning_rate": 7.665094339622642e-05, | |
| "loss": 0.5432, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 5.226676940917969, | |
| "learning_rate": 7.631401617250674e-05, | |
| "loss": 0.557, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 5.462721824645996, | |
| "learning_rate": 7.597708894878706e-05, | |
| "loss": 0.4969, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 4.489198207855225, | |
| "learning_rate": 7.56401617250674e-05, | |
| "loss": 0.5004, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.2848, | |
| "grad_norm": 4.706643104553223, | |
| "learning_rate": 7.530323450134771e-05, | |
| "loss": 0.4465, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 7.614391803741455, | |
| "learning_rate": 7.496630727762804e-05, | |
| "loss": 0.4989, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2912, | |
| "grad_norm": 3.3230834007263184, | |
| "learning_rate": 7.462938005390835e-05, | |
| "loss": 0.4978, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 4.60692834854126, | |
| "learning_rate": 7.429245283018868e-05, | |
| "loss": 0.4477, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 4.93078088760376, | |
| "learning_rate": 7.395552560646901e-05, | |
| "loss": 0.4573, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 3.973932981491089, | |
| "learning_rate": 7.361859838274933e-05, | |
| "loss": 0.4262, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 3.4613964557647705, | |
| "learning_rate": 7.328167115902966e-05, | |
| "loss": 0.4804, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 3.4361777305603027, | |
| "learning_rate": 7.294474393530997e-05, | |
| "loss": 0.4037, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.3104, | |
| "grad_norm": 3.738327980041504, | |
| "learning_rate": 7.26078167115903e-05, | |
| "loss": 0.4436, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 5.539529323577881, | |
| "learning_rate": 7.227088948787062e-05, | |
| "loss": 0.4489, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 3.828397512435913, | |
| "learning_rate": 7.193396226415095e-05, | |
| "loss": 0.4346, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 4.032661437988281, | |
| "learning_rate": 7.159703504043128e-05, | |
| "loss": 0.4615, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3232, | |
| "grad_norm": 3.523750066757202, | |
| "learning_rate": 7.126010781671159e-05, | |
| "loss": 0.4137, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 5.352044582366943, | |
| "learning_rate": 7.092318059299192e-05, | |
| "loss": 0.4397, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.3296, | |
| "grad_norm": 3.491257667541504, | |
| "learning_rate": 7.058625336927224e-05, | |
| "loss": 0.4411, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 2.7892098426818848, | |
| "learning_rate": 7.024932614555257e-05, | |
| "loss": 0.3997, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 3.231407880783081, | |
| "learning_rate": 6.99123989218329e-05, | |
| "loss": 0.4118, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 4.61359977722168, | |
| "learning_rate": 6.957547169811321e-05, | |
| "loss": 0.3631, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.3424, | |
| "grad_norm": 7.804078578948975, | |
| "learning_rate": 6.923854447439353e-05, | |
| "loss": 0.399, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 5.450227737426758, | |
| "learning_rate": 6.890161725067386e-05, | |
| "loss": 0.4628, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.3488, | |
| "grad_norm": 4.55374002456665, | |
| "learning_rate": 6.856469002695418e-05, | |
| "loss": 0.4149, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 2.5150911808013916, | |
| "learning_rate": 6.822776280323451e-05, | |
| "loss": 0.3892, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 3.4889912605285645, | |
| "learning_rate": 6.789083557951483e-05, | |
| "loss": 0.3874, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 4.045688629150391, | |
| "learning_rate": 6.755390835579514e-05, | |
| "loss": 0.4126, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.3616, | |
| "grad_norm": 5.061463356018066, | |
| "learning_rate": 6.721698113207547e-05, | |
| "loss": 0.3564, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 3.681854248046875, | |
| "learning_rate": 6.68800539083558e-05, | |
| "loss": 0.4036, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 4.347032070159912, | |
| "learning_rate": 6.654312668463612e-05, | |
| "loss": 0.4049, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 7.9803314208984375, | |
| "learning_rate": 6.620619946091643e-05, | |
| "loss": 0.3708, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 5.374613285064697, | |
| "learning_rate": 6.586927223719676e-05, | |
| "loss": 0.376, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 3.0691614151000977, | |
| "learning_rate": 6.553234501347709e-05, | |
| "loss": 0.371, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.3808, | |
| "grad_norm": 4.836121082305908, | |
| "learning_rate": 6.519541778975742e-05, | |
| "loss": 0.3681, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 3.778252124786377, | |
| "learning_rate": 6.485849056603774e-05, | |
| "loss": 0.3379, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.3872, | |
| "grad_norm": 3.3291821479797363, | |
| "learning_rate": 6.452156334231805e-05, | |
| "loss": 0.3627, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 3.522620916366577, | |
| "learning_rate": 6.418463611859838e-05, | |
| "loss": 0.3423, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 3.283186912536621, | |
| "learning_rate": 6.384770889487871e-05, | |
| "loss": 0.3301, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 4.433807849884033, | |
| "learning_rate": 6.351078167115904e-05, | |
| "loss": 0.3468, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.97624135017395, | |
| "learning_rate": 6.317385444743936e-05, | |
| "loss": 0.3376, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 3.079582929611206, | |
| "learning_rate": 6.283692722371967e-05, | |
| "loss": 0.3032, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.4064, | |
| "grad_norm": 4.162789821624756, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.3452, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 4.1800665855407715, | |
| "learning_rate": 6.216307277628033e-05, | |
| "loss": 0.3351, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 4.206212520599365, | |
| "learning_rate": 6.182614555256066e-05, | |
| "loss": 0.3579, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 3.900646924972534, | |
| "learning_rate": 6.148921832884098e-05, | |
| "loss": 0.3328, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.4192, | |
| "grad_norm": 5.285961627960205, | |
| "learning_rate": 6.115229110512129e-05, | |
| "loss": 0.345, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 2.616771697998047, | |
| "learning_rate": 6.081536388140162e-05, | |
| "loss": 0.321, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.4256, | |
| "grad_norm": 3.762444496154785, | |
| "learning_rate": 6.047843665768195e-05, | |
| "loss": 0.3188, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 3.244722366333008, | |
| "learning_rate": 6.0141509433962265e-05, | |
| "loss": 0.3299, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 3.7463319301605225, | |
| "learning_rate": 5.980458221024259e-05, | |
| "loss": 0.3063, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 2.2992026805877686, | |
| "learning_rate": 5.9467654986522916e-05, | |
| "loss": 0.3218, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.4384, | |
| "grad_norm": 3.3672423362731934, | |
| "learning_rate": 5.913072776280324e-05, | |
| "loss": 0.3295, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 3.5933585166931152, | |
| "learning_rate": 5.879380053908357e-05, | |
| "loss": 0.282, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.4448, | |
| "grad_norm": 5.269179821014404, | |
| "learning_rate": 5.8456873315363884e-05, | |
| "loss": 0.3113, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 2.99794340133667, | |
| "learning_rate": 5.8119946091644206e-05, | |
| "loss": 0.2977, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 3.103388547897339, | |
| "learning_rate": 5.7783018867924535e-05, | |
| "loss": 0.2954, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 2.7087464332580566, | |
| "learning_rate": 5.744609164420486e-05, | |
| "loss": 0.2779, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.4576, | |
| "grad_norm": 4.316291809082031, | |
| "learning_rate": 5.710916442048517e-05, | |
| "loss": 0.3277, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 3.2305455207824707, | |
| "learning_rate": 5.6772237196765496e-05, | |
| "loss": 0.2738, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 2.637054443359375, | |
| "learning_rate": 5.6435309973045825e-05, | |
| "loss": 0.2764, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 2.727341890335083, | |
| "learning_rate": 5.609838274932615e-05, | |
| "loss": 0.2624, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 3.334686756134033, | |
| "learning_rate": 5.5761455525606476e-05, | |
| "loss": 0.2889, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 2.98366641998291, | |
| "learning_rate": 5.542452830188679e-05, | |
| "loss": 0.3006, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.4768, | |
| "grad_norm": 3.3115627765655518, | |
| "learning_rate": 5.5087601078167114e-05, | |
| "loss": 0.2628, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 4.2003583908081055, | |
| "learning_rate": 5.4750673854447444e-05, | |
| "loss": 0.2607, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4832, | |
| "grad_norm": 6.368921756744385, | |
| "learning_rate": 5.4413746630727766e-05, | |
| "loss": 0.2869, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 6.977168083190918, | |
| "learning_rate": 5.407681940700808e-05, | |
| "loss": 0.2757, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 2.90818452835083, | |
| "learning_rate": 5.373989218328841e-05, | |
| "loss": 0.2545, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 3.735295534133911, | |
| "learning_rate": 5.340296495956873e-05, | |
| "loss": 0.2632, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 3.361546039581299, | |
| "learning_rate": 5.306603773584906e-05, | |
| "loss": 0.2347, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 4.228140830993652, | |
| "learning_rate": 5.2729110512129385e-05, | |
| "loss": 0.2601, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.5024, | |
| "grad_norm": 4.071884632110596, | |
| "learning_rate": 5.23921832884097e-05, | |
| "loss": 0.2604, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 2.7090647220611572, | |
| "learning_rate": 5.205525606469003e-05, | |
| "loss": 0.2353, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 5.145593166351318, | |
| "learning_rate": 5.171832884097035e-05, | |
| "loss": 0.2311, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 3.904247522354126, | |
| "learning_rate": 5.138140161725068e-05, | |
| "loss": 0.2466, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.5152, | |
| "grad_norm": 1.8991984128952026, | |
| "learning_rate": 5.1044474393531e-05, | |
| "loss": 0.267, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 3.3171005249023438, | |
| "learning_rate": 5.070754716981132e-05, | |
| "loss": 0.2648, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.5216, | |
| "grad_norm": 2.8795769214630127, | |
| "learning_rate": 5.037061994609165e-05, | |
| "loss": 0.2293, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 3.4193594455718994, | |
| "learning_rate": 5.003369272237197e-05, | |
| "loss": 0.2243, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 3.470912218093872, | |
| "learning_rate": 4.969676549865229e-05, | |
| "loss": 0.2339, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 4.2241926193237305, | |
| "learning_rate": 4.9359838274932616e-05, | |
| "loss": 0.2346, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.5344, | |
| "grad_norm": 2.4578678607940674, | |
| "learning_rate": 4.902291105121294e-05, | |
| "loss": 0.2162, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 2.8263041973114014, | |
| "learning_rate": 4.868598382749327e-05, | |
| "loss": 0.2254, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.5408, | |
| "grad_norm": 2.4483835697174072, | |
| "learning_rate": 4.834905660377358e-05, | |
| "loss": 0.2099, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 3.059027671813965, | |
| "learning_rate": 4.801212938005391e-05, | |
| "loss": 0.2054, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 2.2340033054351807, | |
| "learning_rate": 4.7675202156334234e-05, | |
| "loss": 0.1919, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 3.059826374053955, | |
| "learning_rate": 4.733827493261456e-05, | |
| "loss": 0.202, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.5536, | |
| "grad_norm": 2.4581353664398193, | |
| "learning_rate": 4.7001347708894886e-05, | |
| "loss": 0.2025, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 3.7301676273345947, | |
| "learning_rate": 4.66644204851752e-05, | |
| "loss": 0.2119, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.9101502895355225, | |
| "learning_rate": 4.632749326145553e-05, | |
| "loss": 0.2102, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.4716150760650635, | |
| "learning_rate": 4.5990566037735846e-05, | |
| "loss": 0.1975, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 1.810396432876587, | |
| "learning_rate": 4.5653638814016176e-05, | |
| "loss": 0.1818, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 2.5587878227233887, | |
| "learning_rate": 4.53167115902965e-05, | |
| "loss": 0.205, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.5728, | |
| "grad_norm": 2.6180355548858643, | |
| "learning_rate": 4.497978436657682e-05, | |
| "loss": 0.2066, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 3.3715500831604004, | |
| "learning_rate": 4.464285714285715e-05, | |
| "loss": 0.1986, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5792, | |
| "grad_norm": 1.9417994022369385, | |
| "learning_rate": 4.4305929919137465e-05, | |
| "loss": 0.18, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 2.6988582611083984, | |
| "learning_rate": 4.3969002695417794e-05, | |
| "loss": 0.1848, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 2.6839194297790527, | |
| "learning_rate": 4.363207547169812e-05, | |
| "loss": 0.2074, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 2.4248201847076416, | |
| "learning_rate": 4.329514824797844e-05, | |
| "loss": 0.1929, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 2.474905252456665, | |
| "learning_rate": 4.295822102425876e-05, | |
| "loss": 0.1761, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 1.7574677467346191, | |
| "learning_rate": 4.2621293800539084e-05, | |
| "loss": 0.1693, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.5984, | |
| "grad_norm": 2.314537286758423, | |
| "learning_rate": 4.2284366576819406e-05, | |
| "loss": 0.1683, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 2.127652406692505, | |
| "learning_rate": 4.1947439353099736e-05, | |
| "loss": 0.2025, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 3.3728537559509277, | |
| "learning_rate": 4.161051212938006e-05, | |
| "loss": 0.1697, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 1.9143751859664917, | |
| "learning_rate": 4.127358490566038e-05, | |
| "loss": 0.1669, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.6112, | |
| "grad_norm": 2.564091205596924, | |
| "learning_rate": 4.09366576819407e-05, | |
| "loss": 0.1695, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 3.54518985748291, | |
| "learning_rate": 4.0599730458221025e-05, | |
| "loss": 0.1836, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.6176, | |
| "grad_norm": 1.8193870782852173, | |
| "learning_rate": 4.026280323450135e-05, | |
| "loss": 0.1701, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 3.027362585067749, | |
| "learning_rate": 3.992587601078167e-05, | |
| "loss": 0.1572, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 2.244178295135498, | |
| "learning_rate": 3.9588948787062e-05, | |
| "loss": 0.164, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 3.8362321853637695, | |
| "learning_rate": 3.9252021563342315e-05, | |
| "loss": 0.1667, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.6304, | |
| "grad_norm": 3.5789124965667725, | |
| "learning_rate": 3.8915094339622644e-05, | |
| "loss": 0.1642, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 3.8518402576446533, | |
| "learning_rate": 3.8578167115902966e-05, | |
| "loss": 0.1615, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.6368, | |
| "grad_norm": 2.248826265335083, | |
| "learning_rate": 3.824123989218329e-05, | |
| "loss": 0.1591, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.504639148712158, | |
| "learning_rate": 3.790431266846362e-05, | |
| "loss": 0.1679, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 2.4511938095092773, | |
| "learning_rate": 3.7567385444743934e-05, | |
| "loss": 0.1887, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 2.2310004234313965, | |
| "learning_rate": 3.723045822102426e-05, | |
| "loss": 0.1584, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.6496, | |
| "grad_norm": 2.4412670135498047, | |
| "learning_rate": 3.689353099730458e-05, | |
| "loss": 0.1493, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 2.1844723224639893, | |
| "learning_rate": 3.655660377358491e-05, | |
| "loss": 0.1636, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 2.738367795944214, | |
| "learning_rate": 3.621967654986524e-05, | |
| "loss": 0.1602, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 2.138552188873291, | |
| "learning_rate": 3.588274932614555e-05, | |
| "loss": 0.159, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 1.3966834545135498, | |
| "learning_rate": 3.554582210242588e-05, | |
| "loss": 0.1428, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 2.3553030490875244, | |
| "learning_rate": 3.52088948787062e-05, | |
| "loss": 0.1542, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.6688, | |
| "grad_norm": 2.553690195083618, | |
| "learning_rate": 3.4871967654986526e-05, | |
| "loss": 0.1433, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 2.555643081665039, | |
| "learning_rate": 3.453504043126685e-05, | |
| "loss": 0.139, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.6752, | |
| "grad_norm": 2.7808148860931396, | |
| "learning_rate": 3.419811320754717e-05, | |
| "loss": 0.1688, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 3.2580206394195557, | |
| "learning_rate": 3.3861185983827494e-05, | |
| "loss": 0.1508, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 4.284587860107422, | |
| "learning_rate": 3.3524258760107816e-05, | |
| "loss": 0.1613, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 1.9140340089797974, | |
| "learning_rate": 3.3187331536388145e-05, | |
| "loss": 0.1562, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 2.13785719871521, | |
| "learning_rate": 3.285040431266847e-05, | |
| "loss": 0.1465, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 2.165259599685669, | |
| "learning_rate": 3.251347708894879e-05, | |
| "loss": 0.1254, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.6944, | |
| "grad_norm": 2.14497971534729, | |
| "learning_rate": 3.217654986522911e-05, | |
| "loss": 0.1436, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 2.4378933906555176, | |
| "learning_rate": 3.1839622641509435e-05, | |
| "loss": 0.154, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 2.225801467895508, | |
| "learning_rate": 3.150269541778976e-05, | |
| "loss": 0.1485, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.9187358617782593, | |
| "learning_rate": 3.1165768194070086e-05, | |
| "loss": 0.1312, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.7072, | |
| "grad_norm": 2.536228895187378, | |
| "learning_rate": 3.08288409703504e-05, | |
| "loss": 0.1574, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 1.6395888328552246, | |
| "learning_rate": 3.0491913746630728e-05, | |
| "loss": 0.133, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.7136, | |
| "grad_norm": 1.9579951763153076, | |
| "learning_rate": 3.0154986522911054e-05, | |
| "loss": 0.139, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 2.6751959323883057, | |
| "learning_rate": 2.9818059299191376e-05, | |
| "loss": 0.1312, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.2726802825927734, | |
| "learning_rate": 2.9481132075471702e-05, | |
| "loss": 0.1324, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 1.9153738021850586, | |
| "learning_rate": 2.914420485175202e-05, | |
| "loss": 0.1271, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.7264, | |
| "grad_norm": 1.5008968114852905, | |
| "learning_rate": 2.8807277628032347e-05, | |
| "loss": 0.1268, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.671263575553894, | |
| "learning_rate": 2.847035040431267e-05, | |
| "loss": 0.1108, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.7328, | |
| "grad_norm": 2.4821507930755615, | |
| "learning_rate": 2.8133423180592995e-05, | |
| "loss": 0.1251, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 1.8863062858581543, | |
| "learning_rate": 2.7796495956873314e-05, | |
| "loss": 0.1224, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 2.204700469970703, | |
| "learning_rate": 2.745956873315364e-05, | |
| "loss": 0.1206, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 3.5700721740722656, | |
| "learning_rate": 2.7122641509433965e-05, | |
| "loss": 0.1379, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.7456, | |
| "grad_norm": 1.4034169912338257, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "loss": 0.1289, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 2.7189619541168213, | |
| "learning_rate": 2.6448787061994614e-05, | |
| "loss": 0.1374, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 1.9794065952301025, | |
| "learning_rate": 2.6111859838274933e-05, | |
| "loss": 0.1306, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.698568344116211, | |
| "learning_rate": 2.577493261455526e-05, | |
| "loss": 0.1279, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 1.7154345512390137, | |
| "learning_rate": 2.5438005390835577e-05, | |
| "loss": 0.1354, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 1.5892575979232788, | |
| "learning_rate": 2.5101078167115903e-05, | |
| "loss": 0.1308, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.7648, | |
| "grad_norm": 1.742138147354126, | |
| "learning_rate": 2.476415094339623e-05, | |
| "loss": 0.1239, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 2.4784200191497803, | |
| "learning_rate": 2.442722371967655e-05, | |
| "loss": 0.1235, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7712, | |
| "grad_norm": 1.3302199840545654, | |
| "learning_rate": 2.4090296495956874e-05, | |
| "loss": 0.1193, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 2.079979658126831, | |
| "learning_rate": 2.3753369272237196e-05, | |
| "loss": 0.1252, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 2.201711893081665, | |
| "learning_rate": 2.341644204851752e-05, | |
| "loss": 0.118, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.476678729057312, | |
| "learning_rate": 2.3079514824797844e-05, | |
| "loss": 0.1116, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 1.3897534608840942, | |
| "learning_rate": 2.274258760107817e-05, | |
| "loss": 0.1116, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 2.085052728652954, | |
| "learning_rate": 2.2405660377358493e-05, | |
| "loss": 0.121, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.7904, | |
| "grad_norm": 1.4823940992355347, | |
| "learning_rate": 2.2068733153638815e-05, | |
| "loss": 0.1242, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 1.9251364469528198, | |
| "learning_rate": 2.1731805929919137e-05, | |
| "loss": 0.1182, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 1.792622447013855, | |
| "learning_rate": 2.1394878706199463e-05, | |
| "loss": 0.1127, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.0481772422790527, | |
| "learning_rate": 2.1057951482479785e-05, | |
| "loss": 0.1137, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8032, | |
| "grad_norm": 2.2373077869415283, | |
| "learning_rate": 2.0721024258760108e-05, | |
| "loss": 0.1162, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.4368358850479126, | |
| "learning_rate": 2.038409703504043e-05, | |
| "loss": 0.1102, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.8096, | |
| "grad_norm": 1.7859045267105103, | |
| "learning_rate": 2.0047169811320756e-05, | |
| "loss": 0.119, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 2.55025315284729, | |
| "learning_rate": 1.971024258760108e-05, | |
| "loss": 0.1168, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 1.7251120805740356, | |
| "learning_rate": 1.9373315363881404e-05, | |
| "loss": 0.1203, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.499519944190979, | |
| "learning_rate": 1.9036388140161727e-05, | |
| "loss": 0.1091, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.8224, | |
| "grad_norm": 2.4695260524749756, | |
| "learning_rate": 1.869946091644205e-05, | |
| "loss": 0.1238, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 1.6461279392242432, | |
| "learning_rate": 1.836253369272237e-05, | |
| "loss": 0.1113, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.8288, | |
| "grad_norm": 2.1009840965270996, | |
| "learning_rate": 1.8025606469002694e-05, | |
| "loss": 0.1153, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.6194204092025757, | |
| "learning_rate": 1.768867924528302e-05, | |
| "loss": 0.1248, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 1.8107424974441528, | |
| "learning_rate": 1.7351752021563345e-05, | |
| "loss": 0.1126, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 1.6274248361587524, | |
| "learning_rate": 1.7014824797843668e-05, | |
| "loss": 0.1119, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.8416, | |
| "grad_norm": 1.5619142055511475, | |
| "learning_rate": 1.667789757412399e-05, | |
| "loss": 0.1121, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 2.297623634338379, | |
| "learning_rate": 1.6340970350404313e-05, | |
| "loss": 0.1246, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 1.5185441970825195, | |
| "learning_rate": 1.600404312668464e-05, | |
| "loss": 0.103, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 1.9803398847579956, | |
| "learning_rate": 1.566711590296496e-05, | |
| "loss": 0.1173, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 2.065286874771118, | |
| "learning_rate": 1.5330188679245283e-05, | |
| "loss": 0.1166, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 1.6951541900634766, | |
| "learning_rate": 1.4993261455525606e-05, | |
| "loss": 0.1036, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.8608, | |
| "grad_norm": 1.2760865688323975, | |
| "learning_rate": 1.465633423180593e-05, | |
| "loss": 0.1086, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 2.6494252681732178, | |
| "learning_rate": 1.4319407008086256e-05, | |
| "loss": 0.1063, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8672, | |
| "grad_norm": 1.4181417226791382, | |
| "learning_rate": 1.3982479784366578e-05, | |
| "loss": 0.1159, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 1.8402090072631836, | |
| "learning_rate": 1.3645552560646902e-05, | |
| "loss": 0.1098, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 2.2500264644622803, | |
| "learning_rate": 1.3308625336927224e-05, | |
| "loss": 0.1126, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 1.485680103302002, | |
| "learning_rate": 1.2971698113207547e-05, | |
| "loss": 0.106, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.3506594896316528, | |
| "learning_rate": 1.2634770889487871e-05, | |
| "loss": 0.1099, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.615838885307312, | |
| "learning_rate": 1.2297843665768195e-05, | |
| "loss": 0.2101, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.8864, | |
| "grad_norm": 1.8304481506347656, | |
| "learning_rate": 1.1960916442048519e-05, | |
| "loss": 0.1022, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 1.7675200700759888, | |
| "learning_rate": 1.1623989218328842e-05, | |
| "loss": 0.102, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 1.363110899925232, | |
| "learning_rate": 1.1287061994609164e-05, | |
| "loss": 0.1078, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 3.771144151687622, | |
| "learning_rate": 1.0950134770889488e-05, | |
| "loss": 0.1099, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8992, | |
| "grad_norm": 1.606154203414917, | |
| "learning_rate": 1.0613207547169812e-05, | |
| "loss": 0.1128, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 1.24965238571167, | |
| "learning_rate": 1.0276280323450135e-05, | |
| "loss": 0.0987, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.9056, | |
| "grad_norm": 1.2326825857162476, | |
| "learning_rate": 9.939353099730459e-06, | |
| "loss": 0.1114, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.5760172605514526, | |
| "learning_rate": 9.602425876010781e-06, | |
| "loss": 0.1114, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 2.7495434284210205, | |
| "learning_rate": 9.265498652291107e-06, | |
| "loss": 0.1108, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 1.2853425741195679, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 0.1015, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.9184, | |
| "grad_norm": 1.6944804191589355, | |
| "learning_rate": 8.591644204851752e-06, | |
| "loss": 0.0976, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.2788165807724, | |
| "learning_rate": 8.254716981132076e-06, | |
| "loss": 0.1088, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.9248, | |
| "grad_norm": 1.9614472389221191, | |
| "learning_rate": 7.9177897574124e-06, | |
| "loss": 0.1057, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 1.3156172037124634, | |
| "learning_rate": 7.580862533692723e-06, | |
| "loss": 0.116, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 2.4010279178619385, | |
| "learning_rate": 7.243935309973046e-06, | |
| "loss": 0.1127, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 2.3485734462738037, | |
| "learning_rate": 6.90700808625337e-06, | |
| "loss": 0.1061, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.9376, | |
| "grad_norm": 2.0748932361602783, | |
| "learning_rate": 6.570080862533692e-06, | |
| "loss": 0.1048, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 1.0390431880950928, | |
| "learning_rate": 6.233153638814016e-06, | |
| "loss": 0.1039, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 1.1470861434936523, | |
| "learning_rate": 5.89622641509434e-06, | |
| "loss": 0.1021, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.8556820154190063, | |
| "learning_rate": 5.5592991913746634e-06, | |
| "loss": 0.1025, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 1.5888488292694092, | |
| "learning_rate": 5.222371967654987e-06, | |
| "loss": 0.1126, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 1.8423320055007935, | |
| "learning_rate": 4.88544474393531e-06, | |
| "loss": 0.1084, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.9568, | |
| "grad_norm": 1.38668954372406, | |
| "learning_rate": 4.548517520215634e-06, | |
| "loss": 0.1029, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.0503017902374268, | |
| "learning_rate": 4.211590296495957e-06, | |
| "loss": 0.1082, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.9632, | |
| "grad_norm": 1.526978850364685, | |
| "learning_rate": 3.8746630727762805e-06, | |
| "loss": 0.0986, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 2.662074089050293, | |
| "learning_rate": 3.5377358490566038e-06, | |
| "loss": 0.1093, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 1.9316941499710083, | |
| "learning_rate": 3.200808625336928e-06, | |
| "loss": 0.0976, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 2.845834732055664, | |
| "learning_rate": 2.8638814016172507e-06, | |
| "loss": 0.1076, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 2.393538475036621, | |
| "learning_rate": 2.5269541778975744e-06, | |
| "loss": 0.1015, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 2.8218283653259277, | |
| "learning_rate": 2.1900269541778976e-06, | |
| "loss": 0.1197, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.9824, | |
| "grad_norm": 1.0874052047729492, | |
| "learning_rate": 1.853099730458221e-06, | |
| "loss": 0.0988, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 1.1840276718139648, | |
| "learning_rate": 1.5161725067385445e-06, | |
| "loss": 0.1024, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 1.8606740236282349, | |
| "learning_rate": 1.179245283018868e-06, | |
| "loss": 0.106, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 1.9203367233276367, | |
| "learning_rate": 8.423180592991913e-07, | |
| "loss": 0.1041, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.9952, | |
| "grad_norm": 1.15613853931427, | |
| "learning_rate": 5.053908355795148e-07, | |
| "loss": 0.114, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 2.514665365219116, | |
| "learning_rate": 1.684636118598383e-07, | |
| "loss": 0.1012, | |
| "step": 3120 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3125, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |