| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9995202522102666, | |
| "eval_steps": 365, | |
| "global_step": 1823, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005482831882667398, | |
| "grad_norm": 1.8271640439809084e-13, | |
| "learning_rate": 9.999257569732963e-06, | |
| "loss": 0.0547, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010965663765334796, | |
| "grad_norm": 1.8681511551469146e-14, | |
| "learning_rate": 9.997030499412933e-06, | |
| "loss": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016448495648002194, | |
| "grad_norm": 1.2511433729323063e-11, | |
| "learning_rate": 9.993319450417669e-06, | |
| "loss": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.021931327530669592, | |
| "grad_norm": 3.090452546650191e-11, | |
| "learning_rate": 9.988125524825216e-06, | |
| "loss": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02741415941333699, | |
| "grad_norm": 1.458404726736688e-14, | |
| "learning_rate": 9.981450265086598e-06, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03289699129600439, | |
| "grad_norm": 3.242672515051115e-14, | |
| "learning_rate": 9.973295653567764e-06, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.038379823178671786, | |
| "grad_norm": 2.718872761479918e-14, | |
| "learning_rate": 9.963664111960875e-06, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.043862655061339184, | |
| "grad_norm": 9.169036085943105e-10, | |
| "learning_rate": 9.952558500565135e-06, | |
| "loss": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04934548694400658, | |
| "grad_norm": 6.057082835297933e-09, | |
| "learning_rate": 9.939982117437356e-06, | |
| "loss": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05482831882667398, | |
| "grad_norm": 2.746146947174566e-06, | |
| "learning_rate": 9.925938697412536e-06, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06031115070934138, | |
| "grad_norm": 2.1543860384554137e-07, | |
| "learning_rate": 9.910432410994701e-06, | |
| "loss": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06579398259200878, | |
| "grad_norm": 1.2734583378914976e-06, | |
| "learning_rate": 9.893467863118402e-06, | |
| "loss": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07127681447467617, | |
| "grad_norm": 3.238647128256389e-09, | |
| "learning_rate": 9.87505009178116e-06, | |
| "loss": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07675964635734357, | |
| "grad_norm": 1.8630474585279444e-07, | |
| "learning_rate": 9.855184566547333e-06, | |
| "loss": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08224247824001096, | |
| "grad_norm": 1.8918351329944016e-09, | |
| "learning_rate": 9.833877186923803e-06, | |
| "loss": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08772531012267837, | |
| "grad_norm": 1.6541060188224321e-10, | |
| "learning_rate": 9.811134280607988e-06, | |
| "loss": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09320814200534576, | |
| "grad_norm": 0.0067967576906085014, | |
| "learning_rate": 9.786962601608689e-06, | |
| "loss": 0.0015, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.09869097388801316, | |
| "grad_norm": 7.394525164272636e-05, | |
| "learning_rate": 9.761369328240347e-06, | |
| "loss": 0.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.10417380577068056, | |
| "grad_norm": 2.62540652329335e-07, | |
| "learning_rate": 9.734362060991274e-06, | |
| "loss": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10965663765334796, | |
| "grad_norm": 4.609726822479843e-09, | |
| "learning_rate": 9.70594882026652e-06, | |
| "loss": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11513946953601535, | |
| "grad_norm": 8.970529319363152e-10, | |
| "learning_rate": 9.676138044006051e-06, | |
| "loss": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.12062230141868276, | |
| "grad_norm": 7.758627629250725e-10, | |
| "learning_rate": 9.644938585178894e-06, | |
| "loss": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12610513330135015, | |
| "grad_norm": 3.484149357912969e-10, | |
| "learning_rate": 9.61235970915407e-06, | |
| "loss": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13158796518401755, | |
| "grad_norm": 3.5987784974267356e-10, | |
| "learning_rate": 9.57841109094903e-06, | |
| "loss": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13707079706668493, | |
| "grad_norm": 2.9926486289078014e-10, | |
| "learning_rate": 9.543102812356446e-06, | |
| "loss": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.14255362894935233, | |
| "grad_norm": 2.17573001282112e-10, | |
| "learning_rate": 9.506445358950199e-06, | |
| "loss": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14803646083201974, | |
| "grad_norm": 1.5743054082584251e-10, | |
| "learning_rate": 9.468449616971458e-06, | |
| "loss": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15351929271468714, | |
| "grad_norm": 3.0337881185005244e-07, | |
| "learning_rate": 9.429126870095766e-06, | |
| "loss": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.15900212459735452, | |
| "grad_norm": 2.075781768695606e-10, | |
| "learning_rate": 9.388488796082113e-06, | |
| "loss": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.16448495648002193, | |
| "grad_norm": 2.6534453922977264e-07, | |
| "learning_rate": 9.346547463304953e-06, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16996778836268933, | |
| "grad_norm": 2.213984316767892e-06, | |
| "learning_rate": 9.303315327170244e-06, | |
| "loss": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.17545062024535674, | |
| "grad_norm": 1.0965933022744068e-10, | |
| "learning_rate": 9.258805226416536e-06, | |
| "loss": 0.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1809334521280241, | |
| "grad_norm": 9.682482593476038e-11, | |
| "learning_rate": 9.213030379302223e-06, | |
| "loss": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18641628401069152, | |
| "grad_norm": 1.8911376631347565e-10, | |
| "learning_rate": 9.166004379680094e-06, | |
| "loss": 0.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.19189911589335892, | |
| "grad_norm": 2.287967731939844e-10, | |
| "learning_rate": 9.11774119296033e-06, | |
| "loss": 0.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19738194777602633, | |
| "grad_norm": 1.6538861946635564e-10, | |
| "learning_rate": 9.068255151963174e-06, | |
| "loss": 0.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.20012336371736, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 8.152250430271124e-09, | |
| "eval_runtime": 172.1258, | |
| "eval_samples_per_second": 37.455, | |
| "eval_steps_per_second": 4.683, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2028647796586937, | |
| "grad_norm": 1.5171223139986978e-06, | |
| "learning_rate": 9.017560952662478e-06, | |
| "loss": 0.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2083476115413611, | |
| "grad_norm": 1.40862849162815e-10, | |
| "learning_rate": 8.965673649821411e-06, | |
| "loss": 0.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.21383044342402852, | |
| "grad_norm": 1.6394713364675795e-10, | |
| "learning_rate": 8.912608652521618e-06, | |
| "loss": 0.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21931327530669592, | |
| "grad_norm": 9.934635752939514e-11, | |
| "learning_rate": 8.858381719587143e-06, | |
| "loss": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2247961071893633, | |
| "grad_norm": 1.0912845627153445e-10, | |
| "learning_rate": 8.803008954904504e-06, | |
| "loss": 0.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2302789390720307, | |
| "grad_norm": 4.275700121070258e-05, | |
| "learning_rate": 8.74650680264029e-06, | |
| "loss": 0.0001, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2357617709546981, | |
| "grad_norm": 3.5764787753578275e-05, | |
| "learning_rate": 8.688892042357703e-06, | |
| "loss": 0.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2412446028373655, | |
| "grad_norm": 8.193592657335103e-05, | |
| "learning_rate": 8.63018178403348e-06, | |
| "loss": 0.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.2467274347200329, | |
| "grad_norm": 9.473233149037696e-06, | |
| "learning_rate": 8.570393462976731e-06, | |
| "loss": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2522102666027003, | |
| "grad_norm": 0.00011029987945221364, | |
| "learning_rate": 8.509544834651124e-06, | |
| "loss": 0.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2576930984853677, | |
| "grad_norm": 6.70829012960894e-06, | |
| "learning_rate": 8.44765396940201e-06, | |
| "loss": 0.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2631759303680351, | |
| "grad_norm": 6.388377641997067e-06, | |
| "learning_rate": 8.384739247090032e-06, | |
| "loss": 0.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2686587622507025, | |
| "grad_norm": 5.076117304270156e-06, | |
| "learning_rate": 8.320819351632828e-06, | |
| "loss": 0.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.27414159413336986, | |
| "grad_norm": 8.84547353052767e-06, | |
| "learning_rate": 8.255913265456416e-06, | |
| "loss": 0.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27962442601603726, | |
| "grad_norm": 3.398100943741156e-06, | |
| "learning_rate": 8.190040263857955e-06, | |
| "loss": 0.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.28510725789870467, | |
| "grad_norm": 8.591404366597999e-06, | |
| "learning_rate": 8.123219909281511e-06, | |
| "loss": 0.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2905900897813721, | |
| "grad_norm": 3.0375645110325422e-06, | |
| "learning_rate": 8.05547204550856e-06, | |
| "loss": 0.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2960729216640395, | |
| "grad_norm": 3.1508220672549214e-06, | |
| "learning_rate": 7.986816791764942e-06, | |
| "loss": 0.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.3015557535467069, | |
| "grad_norm": 4.118597907165622e-09, | |
| "learning_rate": 7.917274536746002e-06, | |
| "loss": 0.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3070385854293743, | |
| "grad_norm": 6.056056697789636e-12, | |
| "learning_rate": 7.846865932561727e-06, | |
| "loss": 0.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3125214173120417, | |
| "grad_norm": 1.4811193160824132e-08, | |
| "learning_rate": 7.775611888603641e-06, | |
| "loss": 0.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.31800424919470904, | |
| "grad_norm": 1.5655763144217794e-11, | |
| "learning_rate": 7.703533565335293e-06, | |
| "loss": 0.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.32348708107737645, | |
| "grad_norm": 2.8082117187189226e-10, | |
| "learning_rate": 7.630652368008202e-06, | |
| "loss": 0.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.32896991296004385, | |
| "grad_norm": 1.598878668351844e-10, | |
| "learning_rate": 7.556989940305087e-06, | |
| "loss": 0.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.33445274484271126, | |
| "grad_norm": 2.0122902455455005e-08, | |
| "learning_rate": 7.482568157912293e-06, | |
| "loss": 0.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33993557672537866, | |
| "grad_norm": 4.7687648101202296e-12, | |
| "learning_rate": 7.407409122023334e-06, | |
| "loss": 0.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.34541840860804607, | |
| "grad_norm": 4.5280151156346093e-11, | |
| "learning_rate": 7.3315351527754405e-06, | |
| "loss": 0.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3509012404907135, | |
| "grad_norm": 1.5659229291031806e-07, | |
| "learning_rate": 7.254968782621115e-06, | |
| "loss": 0.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3563840723733809, | |
| "grad_norm": 1.5579940115806323e-11, | |
| "learning_rate": 7.177732749636612e-06, | |
| "loss": 0.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3618669042560482, | |
| "grad_norm": 9.450333493532526e-08, | |
| "learning_rate": 7.099849990769371e-06, | |
| "loss": 0.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.36734973613871563, | |
| "grad_norm": 4.771372630330006e-09, | |
| "learning_rate": 7.021343635026373e-06, | |
| "loss": 0.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.37283256802138304, | |
| "grad_norm": 9.640462733662147e-11, | |
| "learning_rate": 6.9422369966054856e-06, | |
| "loss": 0.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.37831539990405044, | |
| "grad_norm": 5.622925458581562e-10, | |
| "learning_rate": 6.862553567971782e-06, | |
| "loss": 0.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.38379823178671785, | |
| "grad_norm": 1.4773490590591631e-12, | |
| "learning_rate": 6.782317012880941e-06, | |
| "loss": 0.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.38928106366938525, | |
| "grad_norm": 1.792624208796223e-11, | |
| "learning_rate": 6.7015511593517735e-06, | |
| "loss": 0.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.39476389555205266, | |
| "grad_norm": 1.1963580126561624e-10, | |
| "learning_rate": 6.6202799925899595e-06, | |
| "loss": 0.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.40024672743472, | |
| "grad_norm": 4.0963617047029555e-12, | |
| "learning_rate": 6.5385276478651164e-06, | |
| "loss": 0.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.40024672743472, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 1.5134760111834566e-10, | |
| "eval_runtime": 172.6946, | |
| "eval_samples_per_second": 37.332, | |
| "eval_steps_per_second": 4.667, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.4057295593173874, | |
| "grad_norm": 1.4070650027064868e-11, | |
| "learning_rate": 6.456318403343291e-06, | |
| "loss": 0.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.4112123912000548, | |
| "grad_norm": 1.0772934627700792e-10, | |
| "learning_rate": 6.373676672877034e-06, | |
| "loss": 0.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4166952230827222, | |
| "grad_norm": 3.4152651817720425e-09, | |
| "learning_rate": 6.2906269987551505e-06, | |
| "loss": 0.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4221780549653896, | |
| "grad_norm": 2.17619100517652e-09, | |
| "learning_rate": 6.207194044414333e-06, | |
| "loss": 0.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.42766088684805703, | |
| "grad_norm": 1.4359431612515072e-12, | |
| "learning_rate": 6.123402587114811e-06, | |
| "loss": 0.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.43314371873072444, | |
| "grad_norm": 1.1758019417048432e-13, | |
| "learning_rate": 6.0392775105821936e-06, | |
| "loss": 0.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.43862655061339184, | |
| "grad_norm": 2.0359106067679988e-10, | |
| "learning_rate": 5.954843797617691e-06, | |
| "loss": 0.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4441093824960592, | |
| "grad_norm": 5.402939762255698e-10, | |
| "learning_rate": 5.870126522678933e-06, | |
| "loss": 0.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4495922143787266, | |
| "grad_norm": 4.365914921256042e-11, | |
| "learning_rate": 5.785150844433537e-06, | |
| "loss": 0.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.455075046261394, | |
| "grad_norm": 6.805337001390688e-13, | |
| "learning_rate": 5.699941998287703e-06, | |
| "loss": 0.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4605578781440614, | |
| "grad_norm": 2.589694138386417e-13, | |
| "learning_rate": 5.614525288891989e-06, | |
| "loss": 0.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4660407100267288, | |
| "grad_norm": 2.5609280029925685e-08, | |
| "learning_rate": 5.528926082626542e-06, | |
| "loss": 0.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4715235419093962, | |
| "grad_norm": 1.1285481013242893e-12, | |
| "learning_rate": 5.443169800067987e-06, | |
| "loss": 0.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4770063737920636, | |
| "grad_norm": 1.6540750713556207e-10, | |
| "learning_rate": 5.3572819084402305e-06, | |
| "loss": 0.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.482489205674731, | |
| "grad_norm": 1.1919898401657747e-10, | |
| "learning_rate": 5.271287914051397e-06, | |
| "loss": 0.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4879720375573984, | |
| "grad_norm": 2.92970747488297e-12, | |
| "learning_rate": 5.185213354719173e-06, | |
| "loss": 0.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4934548694400658, | |
| "grad_norm": 2.3761615938755476e-11, | |
| "learning_rate": 5.09908379218679e-06, | |
| "loss": 0.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4989377013227332, | |
| "grad_norm": 4.64755833817776e-09, | |
| "learning_rate": 5.012924804531889e-06, | |
| "loss": 0.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5044205332054006, | |
| "grad_norm": 3.9917269489819773e-10, | |
| "learning_rate": 4.926761978570556e-06, | |
| "loss": 0.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.509903365088068, | |
| "grad_norm": 2.794019737795139e-10, | |
| "learning_rate": 4.840620902258745e-06, | |
| "loss": 0.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5153861969707354, | |
| "grad_norm": 5.130761451255239e-07, | |
| "learning_rate": 4.754527157093374e-06, | |
| "loss": 0.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5208690288534028, | |
| "grad_norm": 1.3340023530261647e-11, | |
| "learning_rate": 4.668506310515325e-06, | |
| "loss": 0.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5263518607360702, | |
| "grad_norm": 1.227000584469451e-10, | |
| "learning_rate": 4.582583908316639e-06, | |
| "loss": 0.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5318346926187376, | |
| "grad_norm": 1.3804829675223562e-12, | |
| "learning_rate": 4.496785467054119e-06, | |
| "loss": 0.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.537317524501405, | |
| "grad_norm": 1.4683420495112398e-12, | |
| "learning_rate": 4.411136466471628e-06, | |
| "loss": 0.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5428003563840724, | |
| "grad_norm": 1.9661270389548235e-09, | |
| "learning_rate": 4.325662341933316e-06, | |
| "loss": 0.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5482831882667397, | |
| "grad_norm": 2.592374981347234e-09, | |
| "learning_rate": 4.240388476870025e-06, | |
| "loss": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5537660201494071, | |
| "grad_norm": 1.8693929054475894e-13, | |
| "learning_rate": 4.155340195241118e-06, | |
| "loss": 0.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5592488520320745, | |
| "grad_norm": 2.1399457761361873e-09, | |
| "learning_rate": 4.070542754013973e-06, | |
| "loss": 0.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5647316839147419, | |
| "grad_norm": 8.234736781354002e-10, | |
| "learning_rate": 3.98602133566336e-06, | |
| "loss": 0.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5702145157974093, | |
| "grad_norm": 6.851032761934217e-12, | |
| "learning_rate": 3.901801040692963e-06, | |
| "loss": 0.0, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5756973476800767, | |
| "grad_norm": 2.0872885364564553e-10, | |
| "learning_rate": 3.817906880181213e-06, | |
| "loss": 0.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5811801795627441, | |
| "grad_norm": 4.878079717940409e-09, | |
| "learning_rate": 3.734363768353707e-06, | |
| "loss": 0.0, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5866630114454116, | |
| "grad_norm": 1.1738454652743258e-10, | |
| "learning_rate": 3.6511965151843754e-06, | |
| "loss": 0.0, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.592145843328079, | |
| "grad_norm": 1.119752957023934e-09, | |
| "learning_rate": 3.568429819027609e-06, | |
| "loss": 0.0, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5976286752107464, | |
| "grad_norm": 8.4299603031468e-07, | |
| "learning_rate": 3.4860882592835386e-06, | |
| "loss": 0.0, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.6003700911520801, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 1.301232038786182e-10, | |
| "eval_runtime": 170.5676, | |
| "eval_samples_per_second": 37.797, | |
| "eval_steps_per_second": 4.725, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.6031115070934138, | |
| "grad_norm": 2.746873475206746e-12, | |
| "learning_rate": 3.4041962890986395e-06, | |
| "loss": 0.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6085943389760812, | |
| "grad_norm": 6.488060089182568e-12, | |
| "learning_rate": 3.3227782281038322e-06, | |
| "loss": 0.0, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.6140771708587486, | |
| "grad_norm": 2.261536069525505e-09, | |
| "learning_rate": 3.2418582551922207e-06, | |
| "loss": 0.0, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.619560002741416, | |
| "grad_norm": 8.609309931628673e-10, | |
| "learning_rate": 3.161460401338644e-06, | |
| "loss": 0.0, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6250428346240834, | |
| "grad_norm": 1.8204505611296717e-11, | |
| "learning_rate": 3.0816085424631446e-06, | |
| "loss": 0.0, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6305256665067507, | |
| "grad_norm": 3.812865648211705e-11, | |
| "learning_rate": 3.002326392340486e-06, | |
| "loss": 0.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6360084983894181, | |
| "grad_norm": 3.348888278232459e-13, | |
| "learning_rate": 2.9236374955578216e-06, | |
| "loss": 0.0, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6414913302720855, | |
| "grad_norm": 9.471730777477205e-08, | |
| "learning_rate": 2.8455652205226146e-06, | |
| "loss": 0.0, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6469741621547529, | |
| "grad_norm": 1.2743735103640241e-12, | |
| "learning_rate": 2.7681327525228633e-06, | |
| "loss": 0.0, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6524569940374203, | |
| "grad_norm": 1.312070541104049e-08, | |
| "learning_rate": 2.6913630868417275e-06, | |
| "loss": 0.0, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6579398259200877, | |
| "grad_norm": 1.1591673801381042e-11, | |
| "learning_rate": 2.615279021928564e-06, | |
| "loss": 0.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6634226578027551, | |
| "grad_norm": 1.2137901794773143e-10, | |
| "learning_rate": 2.539903152628423e-06, | |
| "loss": 0.0, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6689054896854225, | |
| "grad_norm": 2.375248886465897e-10, | |
| "learning_rate": 2.4652578634720147e-06, | |
| "loss": 0.0, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6743883215680899, | |
| "grad_norm": 1.3065516446530978e-09, | |
| "learning_rate": 2.3913653220281256e-06, | |
| "loss": 0.0, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6798711534507573, | |
| "grad_norm": 1.0660168060783182e-11, | |
| "learning_rate": 2.3182474723204634e-06, | |
| "loss": 0.0003, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6853539853334247, | |
| "grad_norm": 1.7096091209367614e-09, | |
| "learning_rate": 2.2459260283109025e-06, | |
| "loss": 0.0, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6908368172160921, | |
| "grad_norm": 1.6090853094397062e-08, | |
| "learning_rate": 2.1744224674510384e-06, | |
| "loss": 0.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6963196490987595, | |
| "grad_norm": 3.779675683546202e-09, | |
| "learning_rate": 2.1037580243039864e-06, | |
| "loss": 0.0, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.701802480981427, | |
| "grad_norm": 8.770736137364565e-09, | |
| "learning_rate": 2.0339536842383013e-06, | |
| "loss": 0.0, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.7072853128640944, | |
| "grad_norm": 7.484961095372e-09, | |
| "learning_rate": 1.9650301771959203e-06, | |
| "loss": 0.0, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.7127681447467618, | |
| "grad_norm": 1.883010858350076e-09, | |
| "learning_rate": 1.8970079715359358e-06, | |
| "loss": 0.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.718250976629429, | |
| "grad_norm": 1.3645068408507655e-09, | |
| "learning_rate": 1.8299072679560764e-06, | |
| "loss": 0.0, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7237338085120965, | |
| "grad_norm": 7.935481605159111e-09, | |
| "learning_rate": 1.7637479934936492e-06, | |
| "loss": 0.0, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7292166403947639, | |
| "grad_norm": 3.995001662815412e-09, | |
| "learning_rate": 1.698549795607779e-06, | |
| "loss": 0.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7346994722774313, | |
| "grad_norm": 3.9588501921095087e-10, | |
| "learning_rate": 1.634332036344652e-06, | |
| "loss": 0.0, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7401823041600987, | |
| "grad_norm": 7.819238589945599e-08, | |
| "learning_rate": 1.571113786587531e-06, | |
| "loss": 0.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7456651360427661, | |
| "grad_norm": 6.14655548680787e-10, | |
| "learning_rate": 1.508913820393238e-06, | |
| "loss": 0.0, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7511479679254335, | |
| "grad_norm": 8.140911234022497e-08, | |
| "learning_rate": 1.4477506094167743e-06, | |
| "loss": 0.0, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7566307998081009, | |
| "grad_norm": 5.628587596007151e-10, | |
| "learning_rate": 1.387642317425766e-06, | |
| "loss": 0.0, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7621136316907683, | |
| "grad_norm": 9.740036555072606e-11, | |
| "learning_rate": 1.328606794906319e-06, | |
| "loss": 0.0, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.7675964635734357, | |
| "grad_norm": 1.1645804676163607e-07, | |
| "learning_rate": 1.270661573761936e-06, | |
| "loss": 0.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7730792954561031, | |
| "grad_norm": 7.708293447983294e-10, | |
| "learning_rate": 1.2138238621070187e-06, | |
| "loss": 0.0, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7785621273387705, | |
| "grad_norm": 3.555227889862067e-09, | |
| "learning_rate": 1.1581105391565456e-06, | |
| "loss": 0.0, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7840449592214379, | |
| "grad_norm": 5.142852810280374e-09, | |
| "learning_rate": 1.1035381502134085e-06, | |
| "loss": 0.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7895277911041053, | |
| "grad_norm": 2.3604476706573507e-10, | |
| "learning_rate": 1.0501229017549264e-06, | |
| "loss": 0.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.7950106229867727, | |
| "grad_norm": 5.834436822560463e-10, | |
| "learning_rate": 9.978806566199721e-07, | |
| "loss": 0.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.80049345486944, | |
| "grad_norm": 5.675169778562861e-10, | |
| "learning_rate": 9.468269292981458e-07, | |
| "loss": 0.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.80049345486944, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 2.4157550959635898e-11, | |
| "eval_runtime": 172.6444, | |
| "eval_samples_per_second": 37.343, | |
| "eval_steps_per_second": 4.669, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.8059762867521074, | |
| "grad_norm": 6.742395974868032e-09, | |
| "learning_rate": 8.969768813224128e-07, | |
| "loss": 0.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.8114591186347748, | |
| "grad_norm": 1.1905292029990022e-10, | |
| "learning_rate": 8.483453167665412e-07, | |
| "loss": 0.0, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.8169419505174422, | |
| "grad_norm": 2.7659869950014127e-09, | |
| "learning_rate": 8.009466778487207e-07, | |
| "loss": 0.0, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8224247824001096, | |
| "grad_norm": 4.819477705808595e-09, | |
| "learning_rate": 7.547950406426069e-07, | |
| "loss": 0.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.827907614282777, | |
| "grad_norm": 4.073986925590134e-09, | |
| "learning_rate": 7.099041108971366e-07, | |
| "loss": 0.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8333904461654444, | |
| "grad_norm": 1.316788900140864e-09, | |
| "learning_rate": 6.662872199662901e-07, | |
| "loss": 0.0, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8388732780481118, | |
| "grad_norm": 1.0290546192948113e-09, | |
| "learning_rate": 6.239573208500626e-07, | |
| "loss": 0.0, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8443561099307793, | |
| "grad_norm": 1.4087077282454175e-08, | |
| "learning_rate": 5.829269843477736e-07, | |
| "loss": 0.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8498389418134467, | |
| "grad_norm": 4.590867408449384e-10, | |
| "learning_rate": 5.432083953248984e-07, | |
| "loss": 0.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8553217736961141, | |
| "grad_norm": 3.087105682197944e-08, | |
| "learning_rate": 5.048133490944995e-07, | |
| "loss": 0.0, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8608046055787815, | |
| "grad_norm": 3.0670661566034596e-09, | |
| "learning_rate": 4.6775324791434704e-07, | |
| "loss": 0.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8662874374614489, | |
| "grad_norm": 3.477520937877898e-09, | |
| "learning_rate": 4.320390976007677e-07, | |
| "loss": 0.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8717702693441163, | |
| "grad_norm": 1.1442693320651287e-10, | |
| "learning_rate": 3.976815042602211e-07, | |
| "loss": 0.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8772531012267837, | |
| "grad_norm": 4.2184575832493465e-09, | |
| "learning_rate": 3.646906711395898e-07, | |
| "loss": 0.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.882735933109451, | |
| "grad_norm": 5.359482857514308e-10, | |
| "learning_rate": 3.3307639559608807e-07, | |
| "loss": 0.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8882187649921184, | |
| "grad_norm": 3.928372627193255e-10, | |
| "learning_rate": 3.02848066187732e-07, | |
| "loss": 0.0, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8937015968747858, | |
| "grad_norm": 8.331213763312917e-08, | |
| "learning_rate": 2.7401465988518894e-07, | |
| "loss": 0.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8991844287574532, | |
| "grad_norm": 9.844909776290933e-09, | |
| "learning_rate": 2.465847394058779e-07, | |
| "loss": 0.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.9046672606401206, | |
| "grad_norm": 2.0618541596295614e-10, | |
| "learning_rate": 2.2056645067107197e-07, | |
| "loss": 0.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.910150092522788, | |
| "grad_norm": 8.006903473578575e-10, | |
| "learning_rate": 1.9596752038679223e-07, | |
| "loss": 0.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.9156329244054554, | |
| "grad_norm": 3.338698761901071e-10, | |
| "learning_rate": 1.727952537491917e-07, | |
| "loss": 0.0, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.9211157562881228, | |
| "grad_norm": 2.3249064895480842e-09, | |
| "learning_rate": 1.5105653227511286e-07, | |
| "loss": 0.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9265985881707902, | |
| "grad_norm": 2.8821448339044764e-08, | |
| "learning_rate": 1.307578117584707e-07, | |
| "loss": 0.0, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.9320814200534576, | |
| "grad_norm": 1.7130765694872707e-09, | |
| "learning_rate": 1.1190512035306356e-07, | |
| "loss": 0.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.937564251936125, | |
| "grad_norm": 1.5305934297771273e-09, | |
| "learning_rate": 9.450405678237685e-08, | |
| "loss": 0.0, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9430470838187924, | |
| "grad_norm": 2.3416841798962196e-09, | |
| "learning_rate": 7.85597886769196e-08, | |
| "loss": 0.0, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9485299157014598, | |
| "grad_norm": 2.2962311163610138e-07, | |
| "learning_rate": 6.407705103958383e-08, | |
| "loss": 0.0, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9540127475841272, | |
| "grad_norm": 3.0744815027183847e-10, | |
| "learning_rate": 5.106014483947708e-08, | |
| "loss": 0.0, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9594955794667946, | |
| "grad_norm": 1.7505968896713853e-09, | |
| "learning_rate": 3.951293573465853e-08, | |
| "loss": 0.0, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.964978411349462, | |
| "grad_norm": 4.694804614513437e-11, | |
| "learning_rate": 2.9438852924142948e-08, | |
| "loss": 0.0, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9704612432321293, | |
| "grad_norm": 2.355823869315543e-10, | |
| "learning_rate": 2.0840888129528002e-08, | |
| "loss": 0.0, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.9759440751147967, | |
| "grad_norm": 1.5761986160711672e-10, | |
| "learning_rate": 1.3721594706532226e-08, | |
| "loss": 0.0, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9814269069974642, | |
| "grad_norm": 3.816250093713336e-10, | |
| "learning_rate": 8.083086886722658e-09, | |
| "loss": 0.0, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.9869097388801316, | |
| "grad_norm": 5.1638181730595534e-08, | |
| "learning_rate": 3.9270391496459835e-09, | |
| "loss": 0.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.992392570762799, | |
| "grad_norm": 2.3921026937046008e-08, | |
| "learning_rate": 1.2546857255546496e-09, | |
| "loss": 0.0, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9978754026454664, | |
| "grad_norm": 1.7885470882106347e-08, | |
| "learning_rate": 6.682022887505302e-11, | |
| "loss": 0.0, | |
| "step": 1820 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1823, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 365, | |
| "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": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |