PlanPhys-Model-Pre-Training-Scaling / multi_student_without_cot /checkpoint-1800 /trainer_state.json
Upload multi_student_without_cot/checkpoint-1800/trainer_state.json with huggingface_hub
25029d9 verified | { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8547008547008547, | |
| "eval_steps": 500, | |
| "global_step": 1800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004748338081671415, | |
| "grad_norm": 20.21827706063475, | |
| "learning_rate": 8.547008547008548e-07, | |
| "loss": 8.145, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00949667616334283, | |
| "grad_norm": 18.24167682921799, | |
| "learning_rate": 1.8043684710351376e-06, | |
| "loss": 8.0172, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.014245014245014245, | |
| "grad_norm": 11.789191102542203, | |
| "learning_rate": 2.7540360873694207e-06, | |
| "loss": 7.3368, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01899335232668566, | |
| "grad_norm": 7.392323142404382, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 6.3843, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.023741690408357077, | |
| "grad_norm": 5.964800407301951, | |
| "learning_rate": 4.653371320037987e-06, | |
| "loss": 5.5837, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02849002849002849, | |
| "grad_norm": 7.120458677127898, | |
| "learning_rate": 5.60303893637227e-06, | |
| "loss": 4.8579, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03323836657169991, | |
| "grad_norm": 8.344343007491538, | |
| "learning_rate": 6.5527065527065525e-06, | |
| "loss": 3.7314, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.03798670465337132, | |
| "grad_norm": 3.1169502061629655, | |
| "learning_rate": 7.5023741690408365e-06, | |
| "loss": 2.5812, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.042735042735042736, | |
| "grad_norm": 2.157035593083223, | |
| "learning_rate": 8.452041785375119e-06, | |
| "loss": 2.0913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.04748338081671415, | |
| "grad_norm": 1.877682319517242, | |
| "learning_rate": 9.401709401709402e-06, | |
| "loss": 1.9562, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.052231718898385564, | |
| "grad_norm": 2.6870548427576004, | |
| "learning_rate": 1.0351377018043685e-05, | |
| "loss": 1.8719, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05698005698005698, | |
| "grad_norm": 4.980475392094001, | |
| "learning_rate": 1.1301044634377968e-05, | |
| "loss": 1.7976, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 8.145965181044149, | |
| "learning_rate": 1.2250712250712251e-05, | |
| "loss": 1.7596, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06647673314339982, | |
| "grad_norm": 10.041185447423977, | |
| "learning_rate": 1.3200379867046534e-05, | |
| "loss": 1.7159, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07122507122507123, | |
| "grad_norm": 10.11622125534401, | |
| "learning_rate": 1.4150047483380819e-05, | |
| "loss": 1.6898, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07597340930674264, | |
| "grad_norm": 8.736601880824086, | |
| "learning_rate": 1.50997150997151e-05, | |
| "loss": 1.6485, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08072174738841406, | |
| "grad_norm": 13.03281077165562, | |
| "learning_rate": 1.604938271604938e-05, | |
| "loss": 1.6175, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 9.908577710414223, | |
| "learning_rate": 1.6999050332383668e-05, | |
| "loss": 1.5976, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09021842355175688, | |
| "grad_norm": 9.237568820174022, | |
| "learning_rate": 1.794871794871795e-05, | |
| "loss": 1.5488, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0949667616334283, | |
| "grad_norm": 4.68415488864487, | |
| "learning_rate": 1.889838556505223e-05, | |
| "loss": 1.5014, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09971509971509972, | |
| "grad_norm": 4.844935709898889, | |
| "learning_rate": 1.9848053181386517e-05, | |
| "loss": 1.4854, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.10446343779677113, | |
| "grad_norm": 4.323562398013845, | |
| "learning_rate": 2.07977207977208e-05, | |
| "loss": 1.4586, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.10921177587844255, | |
| "grad_norm": 3.4385277344372978, | |
| "learning_rate": 2.174738841405508e-05, | |
| "loss": 1.4301, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11396011396011396, | |
| "grad_norm": 3.077116780247069, | |
| "learning_rate": 2.2697056030389367e-05, | |
| "loss": 1.3928, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.11870845204178537, | |
| "grad_norm": 4.302072263978756, | |
| "learning_rate": 2.364672364672365e-05, | |
| "loss": 1.3771, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 3.5795185303541572, | |
| "learning_rate": 2.459639126305793e-05, | |
| "loss": 1.3484, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.1282051282051282, | |
| "grad_norm": 2.071897753499616, | |
| "learning_rate": 2.5546058879392216e-05, | |
| "loss": 1.3107, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.13295346628679963, | |
| "grad_norm": 2.421429428155855, | |
| "learning_rate": 2.64957264957265e-05, | |
| "loss": 1.2881, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.13770180436847104, | |
| "grad_norm": 3.700036964420459, | |
| "learning_rate": 2.744539411206078e-05, | |
| "loss": 1.2667, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.14245014245014245, | |
| "grad_norm": 2.8958286581854944, | |
| "learning_rate": 2.839506172839506e-05, | |
| "loss": 1.247, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14719848053181386, | |
| "grad_norm": 2.054414468223543, | |
| "learning_rate": 2.9344729344729345e-05, | |
| "loss": 1.2217, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.15194681861348527, | |
| "grad_norm": 1.96235600963249, | |
| "learning_rate": 3.029439696106363e-05, | |
| "loss": 1.2, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.15669515669515668, | |
| "grad_norm": 2.731627929284309, | |
| "learning_rate": 3.1244064577397914e-05, | |
| "loss": 1.1817, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.16144349477682812, | |
| "grad_norm": 2.1653241264851637, | |
| "learning_rate": 3.2193732193732194e-05, | |
| "loss": 1.1623, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.16619183285849953, | |
| "grad_norm": 2.519530114010952, | |
| "learning_rate": 3.314339981006648e-05, | |
| "loss": 1.1288, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 2.653996907642351, | |
| "learning_rate": 3.409306742640076e-05, | |
| "loss": 1.1081, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.17568850902184235, | |
| "grad_norm": 2.4977140265839126, | |
| "learning_rate": 3.504273504273504e-05, | |
| "loss": 1.0854, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.18043684710351376, | |
| "grad_norm": 2.856768490699816, | |
| "learning_rate": 3.599240265906933e-05, | |
| "loss": 1.0636, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 2.17649907858301, | |
| "learning_rate": 3.694207027540361e-05, | |
| "loss": 1.0478, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1899335232668566, | |
| "grad_norm": 3.9461445509883593, | |
| "learning_rate": 3.789173789173789e-05, | |
| "loss": 1.0281, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.19468186134852802, | |
| "grad_norm": 2.871728182164183, | |
| "learning_rate": 3.884140550807218e-05, | |
| "loss": 0.9991, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19943019943019943, | |
| "grad_norm": 3.013330420912045, | |
| "learning_rate": 3.979107312440646e-05, | |
| "loss": 0.9773, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.20417853751187084, | |
| "grad_norm": 2.420232752543592, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.9586, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.20892687559354226, | |
| "grad_norm": 2.5768091814664564, | |
| "learning_rate": 4.169040835707503e-05, | |
| "loss": 0.938, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.21367521367521367, | |
| "grad_norm": 3.113768661249082, | |
| "learning_rate": 4.264007597340931e-05, | |
| "loss": 0.9293, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2184235517568851, | |
| "grad_norm": 2.294957751769951, | |
| "learning_rate": 4.358974358974359e-05, | |
| "loss": 0.9049, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.22317188983855651, | |
| "grad_norm": 2.4492691119209855, | |
| "learning_rate": 4.453941120607788e-05, | |
| "loss": 0.8865, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.22792022792022792, | |
| "grad_norm": 2.286177237412566, | |
| "learning_rate": 4.548907882241216e-05, | |
| "loss": 0.866, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.23266856600189934, | |
| "grad_norm": 2.589513226874877, | |
| "learning_rate": 4.643874643874644e-05, | |
| "loss": 0.8457, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.23741690408357075, | |
| "grad_norm": 3.0229688171419875, | |
| "learning_rate": 4.738841405508073e-05, | |
| "loss": 0.8279, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.24216524216524216, | |
| "grad_norm": 2.686293470195147, | |
| "learning_rate": 4.833808167141501e-05, | |
| "loss": 0.8147, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 2.17482173902714, | |
| "learning_rate": 4.928774928774929e-05, | |
| "loss": 0.7885, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.251661918328585, | |
| "grad_norm": 2.8541347160003783, | |
| "learning_rate": 5.0237416904083576e-05, | |
| "loss": 0.7784, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 2.59370867139089, | |
| "learning_rate": 5.118708452041786e-05, | |
| "loss": 0.7673, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2611585944919278, | |
| "grad_norm": 3.1513543951371905, | |
| "learning_rate": 5.213675213675214e-05, | |
| "loss": 0.7479, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.26590693257359926, | |
| "grad_norm": 2.536640604487201, | |
| "learning_rate": 5.308641975308642e-05, | |
| "loss": 0.7253, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.2706552706552707, | |
| "grad_norm": 2.8005849730532213, | |
| "learning_rate": 5.403608736942071e-05, | |
| "loss": 0.7154, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2754036087369421, | |
| "grad_norm": 2.551754718587974, | |
| "learning_rate": 5.498575498575499e-05, | |
| "loss": 0.6901, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.2801519468186135, | |
| "grad_norm": 2.6731387859043365, | |
| "learning_rate": 5.5935422602089274e-05, | |
| "loss": 0.6784, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2849002849002849, | |
| "grad_norm": 2.558888468671125, | |
| "learning_rate": 5.6885090218423554e-05, | |
| "loss": 0.6634, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2896486229819563, | |
| "grad_norm": 2.7304234847433477, | |
| "learning_rate": 5.7834757834757834e-05, | |
| "loss": 0.6528, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.2943969610636277, | |
| "grad_norm": 2.469120865455278, | |
| "learning_rate": 5.878442545109212e-05, | |
| "loss": 0.6357, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.29914529914529914, | |
| "grad_norm": 2.361617168312385, | |
| "learning_rate": 5.97340930674264e-05, | |
| "loss": 0.6184, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.30389363722697055, | |
| "grad_norm": 2.5062066018611544, | |
| "learning_rate": 6.068376068376068e-05, | |
| "loss": 0.6016, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 3.35086067384153, | |
| "learning_rate": 6.163342830009497e-05, | |
| "loss": 0.5978, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.31339031339031337, | |
| "grad_norm": 2.177019547847069, | |
| "learning_rate": 6.258309591642925e-05, | |
| "loss": 0.5774, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3181386514719848, | |
| "grad_norm": 2.6347133971158465, | |
| "learning_rate": 6.353276353276355e-05, | |
| "loss": 0.5683, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.32288698955365625, | |
| "grad_norm": 1.9226802841170278, | |
| "learning_rate": 6.448243114909783e-05, | |
| "loss": 0.5504, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.32763532763532766, | |
| "grad_norm": 1.9462440337757687, | |
| "learning_rate": 6.54320987654321e-05, | |
| "loss": 0.5393, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.33238366571699907, | |
| "grad_norm": 2.64699999142231, | |
| "learning_rate": 6.638176638176638e-05, | |
| "loss": 0.5295, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3371320037986705, | |
| "grad_norm": 2.428317113805517, | |
| "learning_rate": 6.733143399810066e-05, | |
| "loss": 0.5172, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 2.2288280010027854, | |
| "learning_rate": 6.828110161443494e-05, | |
| "loss": 0.5013, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3466286799620133, | |
| "grad_norm": 2.417506087683697, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.494, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.3513770180436847, | |
| "grad_norm": 1.9466452806944854, | |
| "learning_rate": 7.018043684710352e-05, | |
| "loss": 0.4766, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3561253561253561, | |
| "grad_norm": 2.548627924889759, | |
| "learning_rate": 7.11301044634378e-05, | |
| "loss": 0.4684, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.36087369420702753, | |
| "grad_norm": 2.157265574032032, | |
| "learning_rate": 7.207977207977208e-05, | |
| "loss": 0.4606, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.36562203228869894, | |
| "grad_norm": 2.037117411340044, | |
| "learning_rate": 7.302943969610636e-05, | |
| "loss": 0.4485, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 1.8379017170113745, | |
| "learning_rate": 7.397910731244065e-05, | |
| "loss": 0.4338, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.37511870845204176, | |
| "grad_norm": 2.8763875380181734, | |
| "learning_rate": 7.492877492877494e-05, | |
| "loss": 0.4289, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.3798670465337132, | |
| "grad_norm": 1.8515635424892907, | |
| "learning_rate": 7.587844254510922e-05, | |
| "loss": 0.4179, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 1.6717226940996468, | |
| "learning_rate": 7.68281101614435e-05, | |
| "loss": 0.4092, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.38936372269705605, | |
| "grad_norm": 2.081127977320627, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 0.3973, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.39411206077872746, | |
| "grad_norm": 1.8193841823557135, | |
| "learning_rate": 7.872744539411206e-05, | |
| "loss": 0.3905, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.39886039886039887, | |
| "grad_norm": 2.029853671008697, | |
| "learning_rate": 7.967711301044634e-05, | |
| "loss": 0.3829, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4036087369420703, | |
| "grad_norm": 1.7467553830120104, | |
| "learning_rate": 8.062678062678063e-05, | |
| "loss": 0.3705, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4083570750237417, | |
| "grad_norm": 1.8021994967000676, | |
| "learning_rate": 8.157644824311491e-05, | |
| "loss": 0.3639, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4131054131054131, | |
| "grad_norm": 1.641013312643542, | |
| "learning_rate": 8.25261158594492e-05, | |
| "loss": 0.3608, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4178537511870845, | |
| "grad_norm": 1.5079068085978562, | |
| "learning_rate": 8.347578347578347e-05, | |
| "loss": 0.3522, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4226020892687559, | |
| "grad_norm": 1.4655692445142585, | |
| "learning_rate": 8.442545109211775e-05, | |
| "loss": 0.3443, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 1.4792862368111255, | |
| "learning_rate": 8.537511870845205e-05, | |
| "loss": 0.3373, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.43209876543209874, | |
| "grad_norm": 1.7589212642036423, | |
| "learning_rate": 8.632478632478634e-05, | |
| "loss": 0.3349, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4368471035137702, | |
| "grad_norm": 1.3662339866574285, | |
| "learning_rate": 8.727445394112062e-05, | |
| "loss": 0.3264, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.4415954415954416, | |
| "grad_norm": 1.6091099458218114, | |
| "learning_rate": 8.82241215574549e-05, | |
| "loss": 0.3181, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.44634377967711303, | |
| "grad_norm": 1.347407542639701, | |
| "learning_rate": 8.917378917378918e-05, | |
| "loss": 0.3095, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.45109211775878444, | |
| "grad_norm": 1.9418517203504135, | |
| "learning_rate": 9.012345679012346e-05, | |
| "loss": 0.305, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.45584045584045585, | |
| "grad_norm": 1.3400978709493576, | |
| "learning_rate": 9.107312440645774e-05, | |
| "loss": 0.3008, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.46058879392212726, | |
| "grad_norm": 1.2579527847439156, | |
| "learning_rate": 9.202279202279203e-05, | |
| "loss": 0.2953, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.46533713200379867, | |
| "grad_norm": 1.2589304750507126, | |
| "learning_rate": 9.297245963912631e-05, | |
| "loss": 0.2898, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.4700854700854701, | |
| "grad_norm": 1.3117037531826747, | |
| "learning_rate": 9.392212725546059e-05, | |
| "loss": 0.2843, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.4748338081671415, | |
| "grad_norm": 1.3291209884932123, | |
| "learning_rate": 9.487179487179487e-05, | |
| "loss": 0.2815, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4795821462488129, | |
| "grad_norm": 1.209968301390431, | |
| "learning_rate": 9.582146248812915e-05, | |
| "loss": 0.2799, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4843304843304843, | |
| "grad_norm": 1.0056322036392946, | |
| "learning_rate": 9.677113010446344e-05, | |
| "loss": 0.275, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.4890788224121557, | |
| "grad_norm": 1.2702143901456204, | |
| "learning_rate": 9.772079772079774e-05, | |
| "loss": 0.2691, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 1.0226410737222107, | |
| "learning_rate": 9.867046533713202e-05, | |
| "loss": 0.2616, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.4985754985754986, | |
| "grad_norm": 0.9723093234963888, | |
| "learning_rate": 9.96201329534663e-05, | |
| "loss": 0.2579, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.50332383665717, | |
| "grad_norm": 1.3912534326159531, | |
| "learning_rate": 9.999990109907572e-05, | |
| "loss": 0.2551, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5080721747388414, | |
| "grad_norm": 1.1085969512788294, | |
| "learning_rate": 9.999929670595522e-05, | |
| "loss": 0.2531, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 1.0552909890539075, | |
| "learning_rate": 9.99981428713058e-05, | |
| "loss": 0.2483, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5175688509021842, | |
| "grad_norm": 1.1040506601466529, | |
| "learning_rate": 9.999643960780693e-05, | |
| "loss": 0.2423, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5223171889838556, | |
| "grad_norm": 0.9501946984359841, | |
| "learning_rate": 9.999418693417573e-05, | |
| "loss": 0.2376, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5270655270655271, | |
| "grad_norm": 1.022708548136873, | |
| "learning_rate": 9.999138487516681e-05, | |
| "loss": 0.2343, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5318138651471985, | |
| "grad_norm": 1.1751853468010314, | |
| "learning_rate": 9.998803346157195e-05, | |
| "loss": 0.2362, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5365622032288699, | |
| "grad_norm": 0.8927081511146825, | |
| "learning_rate": 9.99841327302198e-05, | |
| "loss": 0.2251, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5413105413105413, | |
| "grad_norm": 0.9732415935122696, | |
| "learning_rate": 9.997968272397542e-05, | |
| "loss": 0.2222, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.5460588793922128, | |
| "grad_norm": 0.9560857920847607, | |
| "learning_rate": 9.997468349173985e-05, | |
| "loss": 0.2189, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.5508072174738842, | |
| "grad_norm": 0.924098382546985, | |
| "learning_rate": 9.99691350884496e-05, | |
| "loss": 0.2179, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.828396782450117, | |
| "learning_rate": 9.996303757507602e-05, | |
| "loss": 0.211, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.560303893637227, | |
| "grad_norm": 0.9333663244449117, | |
| "learning_rate": 9.995639101862453e-05, | |
| "loss": 0.2102, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.5650522317188984, | |
| "grad_norm": 0.9298060626900368, | |
| "learning_rate": 9.994919549213408e-05, | |
| "loss": 0.2074, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5698005698005698, | |
| "grad_norm": 0.7962626517135125, | |
| "learning_rate": 9.99414510746762e-05, | |
| "loss": 0.203, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5745489078822412, | |
| "grad_norm": 0.8580673639697189, | |
| "learning_rate": 9.993315785135416e-05, | |
| "loss": 0.2013, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5792972459639126, | |
| "grad_norm": 0.8048698470169939, | |
| "learning_rate": 9.99243159133021e-05, | |
| "loss": 0.1975, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.584045584045584, | |
| "grad_norm": 0.9257889896229621, | |
| "learning_rate": 9.991492535768392e-05, | |
| "loss": 0.1946, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.5887939221272555, | |
| "grad_norm": 0.7558640985069887, | |
| "learning_rate": 9.99049862876923e-05, | |
| "loss": 0.1925, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.5935422602089269, | |
| "grad_norm": 0.8056728442959841, | |
| "learning_rate": 9.989449881254754e-05, | |
| "loss": 0.1891, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 0.8438539837910541, | |
| "learning_rate": 9.988346304749635e-05, | |
| "loss": 0.1862, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6030389363722697, | |
| "grad_norm": 0.7510354369473087, | |
| "learning_rate": 9.987187911381059e-05, | |
| "loss": 0.1841, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6077872744539411, | |
| "grad_norm": 0.7636415261243265, | |
| "learning_rate": 9.985974713878592e-05, | |
| "loss": 0.1808, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6125356125356125, | |
| "grad_norm": 0.7974385042371772, | |
| "learning_rate": 9.984706725574044e-05, | |
| "loss": 0.1792, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 0.7789431577859948, | |
| "learning_rate": 9.983383960401319e-05, | |
| "loss": 0.1756, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6220322886989553, | |
| "grad_norm": 0.7722426337147819, | |
| "learning_rate": 9.982006432896266e-05, | |
| "loss": 0.176, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6267806267806267, | |
| "grad_norm": 0.7383737481941087, | |
| "learning_rate": 9.980574158196512e-05, | |
| "loss": 0.1723, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.6315289648622981, | |
| "grad_norm": 0.7442324964032874, | |
| "learning_rate": 9.979087152041304e-05, | |
| "loss": 0.1698, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.6362773029439696, | |
| "grad_norm": 0.7497638974160803, | |
| "learning_rate": 9.977545430771332e-05, | |
| "loss": 0.1673, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "grad_norm": 0.8548555257280125, | |
| "learning_rate": 9.975949011328542e-05, | |
| "loss": 0.1636, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.6457739791073125, | |
| "grad_norm": 0.8279981954343867, | |
| "learning_rate": 9.97429791125597e-05, | |
| "loss": 0.1655, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.6505223171889839, | |
| "grad_norm": 0.6959990480791529, | |
| "learning_rate": 9.972592148697528e-05, | |
| "loss": 0.1612, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.6552706552706553, | |
| "grad_norm": 0.828078740940128, | |
| "learning_rate": 9.970831742397815e-05, | |
| "loss": 0.1612, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.6600189933523267, | |
| "grad_norm": 0.77793285129046, | |
| "learning_rate": 9.969016711701908e-05, | |
| "loss": 0.1611, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.6647673314339981, | |
| "grad_norm": 0.7194789618958793, | |
| "learning_rate": 9.967147076555157e-05, | |
| "loss": 0.1579, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6695156695156695, | |
| "grad_norm": 0.7100892012167056, | |
| "learning_rate": 9.965222857502951e-05, | |
| "loss": 0.1564, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.674264007597341, | |
| "grad_norm": 0.7535295766221981, | |
| "learning_rate": 9.963244075690505e-05, | |
| "loss": 0.1519, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.6790123456790124, | |
| "grad_norm": 0.706459243665132, | |
| "learning_rate": 9.961210752862626e-05, | |
| "loss": 0.1548, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 0.6799536660584122, | |
| "learning_rate": 9.959122911363463e-05, | |
| "loss": 0.1489, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.6885090218423552, | |
| "grad_norm": 0.750734170664292, | |
| "learning_rate": 9.956980574136281e-05, | |
| "loss": 0.1492, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6932573599240266, | |
| "grad_norm": 0.7631275022947855, | |
| "learning_rate": 9.954783764723188e-05, | |
| "loss": 0.1469, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.698005698005698, | |
| "grad_norm": 0.7140096944520468, | |
| "learning_rate": 9.952532507264891e-05, | |
| "loss": 0.1456, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7027540360873694, | |
| "grad_norm": 0.6820968891017294, | |
| "learning_rate": 9.950226826500426e-05, | |
| "loss": 0.1459, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7075023741690408, | |
| "grad_norm": 0.6573844442119157, | |
| "learning_rate": 9.947866747766882e-05, | |
| "loss": 0.1432, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.7122507122507122, | |
| "grad_norm": 0.6477315679395246, | |
| "learning_rate": 9.945452296999131e-05, | |
| "loss": 0.1418, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7169990503323836, | |
| "grad_norm": 0.7854073646179602, | |
| "learning_rate": 9.942983500729536e-05, | |
| "loss": 0.1415, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.7217473884140551, | |
| "grad_norm": 0.6563280291041728, | |
| "learning_rate": 9.940460386087667e-05, | |
| "loss": 0.1396, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.7264957264957265, | |
| "grad_norm": 0.6777516445796997, | |
| "learning_rate": 9.937882980799986e-05, | |
| "loss": 0.1374, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.7312440645773979, | |
| "grad_norm": 0.6482850562773214, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 0.1341, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.7359924026590693, | |
| "grad_norm": 0.6807447861160003, | |
| "learning_rate": 9.93256541217576e-05, | |
| "loss": 0.1328, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.6096762595093985, | |
| "learning_rate": 9.929825307273898e-05, | |
| "loss": 0.1323, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.7454890788224121, | |
| "grad_norm": 0.6758283907647111, | |
| "learning_rate": 9.927031028594951e-05, | |
| "loss": 0.1323, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.7502374169040835, | |
| "grad_norm": 0.6661507717312829, | |
| "learning_rate": 9.924182606845204e-05, | |
| "loss": 0.1292, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.7549857549857549, | |
| "grad_norm": 0.7190496153063343, | |
| "learning_rate": 9.921280073325922e-05, | |
| "loss": 0.1288, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.7597340930674265, | |
| "grad_norm": 0.7792128467633825, | |
| "learning_rate": 9.918323459933005e-05, | |
| "loss": 0.1315, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7644824311490979, | |
| "grad_norm": 0.7476110366866813, | |
| "learning_rate": 9.915312799156632e-05, | |
| "loss": 0.127, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.723682326262704, | |
| "learning_rate": 9.91224812408091e-05, | |
| "loss": 0.1252, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.7739791073124407, | |
| "grad_norm": 0.6763342918277632, | |
| "learning_rate": 9.909129468383506e-05, | |
| "loss": 0.1229, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.7787274453941121, | |
| "grad_norm": 0.6756385241786282, | |
| "learning_rate": 9.905956866335286e-05, | |
| "loss": 0.1223, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.7834757834757835, | |
| "grad_norm": 0.7031791508138145, | |
| "learning_rate": 9.90273035279992e-05, | |
| "loss": 0.121, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.7882241215574549, | |
| "grad_norm": 0.6901786661268531, | |
| "learning_rate": 9.899449963233523e-05, | |
| "loss": 0.1218, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.7929724596391263, | |
| "grad_norm": 0.655344108091334, | |
| "learning_rate": 9.896115733684244e-05, | |
| "loss": 0.1186, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.7977207977207977, | |
| "grad_norm": 0.6429227026044089, | |
| "learning_rate": 9.892727700791881e-05, | |
| "loss": 0.119, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.8024691358024691, | |
| "grad_norm": 0.6445284291633588, | |
| "learning_rate": 9.889285901787482e-05, | |
| "loss": 0.1171, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.8072174738841406, | |
| "grad_norm": 0.6599500074332483, | |
| "learning_rate": 9.885790374492921e-05, | |
| "loss": 0.114, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.811965811965812, | |
| "grad_norm": 0.6458817345788224, | |
| "learning_rate": 9.882241157320497e-05, | |
| "loss": 0.1161, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.8167141500474834, | |
| "grad_norm": 0.6510623099667147, | |
| "learning_rate": 9.878638289272507e-05, | |
| "loss": 0.1134, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.8214624881291548, | |
| "grad_norm": 0.711735747979104, | |
| "learning_rate": 9.874981809940808e-05, | |
| "loss": 0.1133, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.8262108262108262, | |
| "grad_norm": 0.6067344669903363, | |
| "learning_rate": 9.871271759506404e-05, | |
| "loss": 0.111, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.8309591642924976, | |
| "grad_norm": 0.6435729954519224, | |
| "learning_rate": 9.86750817873898e-05, | |
| "loss": 0.1101, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.835707502374169, | |
| "grad_norm": 0.6594866421615497, | |
| "learning_rate": 9.863691108996471e-05, | |
| "loss": 0.1102, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.8404558404558404, | |
| "grad_norm": 0.6463645658702524, | |
| "learning_rate": 9.859820592224599e-05, | |
| "loss": 0.1086, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.8452041785375118, | |
| "grad_norm": 0.6384759072033943, | |
| "learning_rate": 9.855896670956415e-05, | |
| "loss": 0.107, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.8499525166191833, | |
| "grad_norm": 0.5925061730095992, | |
| "learning_rate": 9.851919388311832e-05, | |
| "loss": 0.1057, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 0.5958838267885576, | |
| "learning_rate": 9.84788878799715e-05, | |
| "loss": 0.1063, | |
| "step": 1800 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10530, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 1800, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 113494835331072.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |