| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 1210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08264462809917356, | |
| "grad_norm": 9.173369407653809, | |
| "learning_rate": 1.6393442622950818e-05, | |
| "loss": 1.2749, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1652892561983471, | |
| "grad_norm": 3.0318551063537598, | |
| "learning_rate": 3.2786885245901635e-05, | |
| "loss": 0.7164, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24793388429752067, | |
| "grad_norm": 1.8877290487289429, | |
| "learning_rate": 4.918032786885246e-05, | |
| "loss": 0.4175, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3305785123966942, | |
| "grad_norm": 2.459096670150757, | |
| "learning_rate": 6.557377049180327e-05, | |
| "loss": 0.3476, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4132231404958678, | |
| "grad_norm": 1.8838213682174683, | |
| "learning_rate": 8.19672131147541e-05, | |
| "loss": 0.2797, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.49586776859504134, | |
| "grad_norm": 1.053584337234497, | |
| "learning_rate": 9.836065573770493e-05, | |
| "loss": 0.2348, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5785123966942148, | |
| "grad_norm": 2.7973430156707764, | |
| "learning_rate": 9.998486219870769e-05, | |
| "loss": 0.224, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6611570247933884, | |
| "grad_norm": 1.6049909591674805, | |
| "learning_rate": 9.993254576451651e-05, | |
| "loss": 0.2017, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.743801652892562, | |
| "grad_norm": 0.7833113074302673, | |
| "learning_rate": 9.984290291014105e-05, | |
| "loss": 0.2074, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8264462809917356, | |
| "grad_norm": 3.117323160171509, | |
| "learning_rate": 9.971600064692222e-05, | |
| "loss": 0.1826, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.858580470085144, | |
| "learning_rate": 9.955193383898376e-05, | |
| "loss": 0.1674, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9917355371900827, | |
| "grad_norm": 0.7289248108863831, | |
| "learning_rate": 9.935082513231775e-05, | |
| "loss": 0.1584, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0743801652892562, | |
| "grad_norm": 2.072802782058716, | |
| "learning_rate": 9.911282486310213e-05, | |
| "loss": 0.154, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1570247933884297, | |
| "grad_norm": 1.2961859703063965, | |
| "learning_rate": 9.883811094531906e-05, | |
| "loss": 0.1567, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2396694214876034, | |
| "grad_norm": 0.9249287843704224, | |
| "learning_rate": 9.85268887377574e-05, | |
| "loss": 0.1537, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.322314049586777, | |
| "grad_norm": 1.7591155767440796, | |
| "learning_rate": 9.817939089049964e-05, | |
| "loss": 0.1435, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4049586776859504, | |
| "grad_norm": 3.4378185272216797, | |
| "learning_rate": 9.779587717100729e-05, | |
| "loss": 0.1329, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.487603305785124, | |
| "grad_norm": 0.6799660921096802, | |
| "learning_rate": 9.737663426993513e-05, | |
| "loss": 0.1388, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.5702479338842976, | |
| "grad_norm": 0.6710700988769531, | |
| "learning_rate": 9.69219755868194e-05, | |
| "loss": 0.147, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6528925619834711, | |
| "grad_norm": 1.6649482250213623, | |
| "learning_rate": 9.643224099579998e-05, | |
| "loss": 0.1219, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7355371900826446, | |
| "grad_norm": 1.4101303815841675, | |
| "learning_rate": 9.59077965915521e-05, | |
| "loss": 0.1132, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.7230204939842224, | |
| "learning_rate": 9.534903441561693e-05, | |
| "loss": 0.1196, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9008264462809916, | |
| "grad_norm": 2.3801050186157227, | |
| "learning_rate": 9.47563721633361e-05, | |
| "loss": 0.106, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9834710743801653, | |
| "grad_norm": 1.3875033855438232, | |
| "learning_rate": 9.413025287160905e-05, | |
| "loss": 0.1181, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0661157024793386, | |
| "grad_norm": 0.49982619285583496, | |
| "learning_rate": 9.347114458770657e-05, | |
| "loss": 0.0937, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.1487603305785123, | |
| "grad_norm": 1.5207865238189697, | |
| "learning_rate": 9.277954001938818e-05, | |
| "loss": 0.1005, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.231404958677686, | |
| "grad_norm": 0.5808604955673218, | |
| "learning_rate": 9.205595616658495e-05, | |
| "loss": 0.1046, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.3140495867768593, | |
| "grad_norm": 0.5397279262542725, | |
| "learning_rate": 9.1300933934923e-05, | |
| "loss": 0.095, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.396694214876033, | |
| "grad_norm": 2.023026704788208, | |
| "learning_rate": 9.051503773137646e-05, | |
| "loss": 0.1054, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.479338842975207, | |
| "grad_norm": 0.8514553904533386, | |
| "learning_rate": 8.969885504235256e-05, | |
| "loss": 0.1058, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.56198347107438, | |
| "grad_norm": 0.7182636260986328, | |
| "learning_rate": 8.885299599452382e-05, | |
| "loss": 0.1155, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.644628099173554, | |
| "grad_norm": 1.0667051076889038, | |
| "learning_rate": 8.797809289873587e-05, | |
| "loss": 0.1009, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.7455232739448547, | |
| "learning_rate": 8.70747997773317e-05, | |
| "loss": 0.0969, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.809917355371901, | |
| "grad_norm": 0.5841661691665649, | |
| "learning_rate": 8.614379187524592e-05, | |
| "loss": 0.0885, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8925619834710745, | |
| "grad_norm": 0.5531947016716003, | |
| "learning_rate": 8.518576515523424e-05, | |
| "loss": 0.0909, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.975206611570248, | |
| "grad_norm": 0.7314962148666382, | |
| "learning_rate": 8.42014357776155e-05, | |
| "loss": 0.0821, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.0578512396694215, | |
| "grad_norm": 1.354279637336731, | |
| "learning_rate": 8.319153956491568e-05, | |
| "loss": 0.0908, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.1404958677685952, | |
| "grad_norm": 0.621191143989563, | |
| "learning_rate": 8.215683145181312e-05, | |
| "loss": 0.0822, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.2231404958677685, | |
| "grad_norm": 0.6993005871772766, | |
| "learning_rate": 8.109808492079718e-05, | |
| "loss": 0.0791, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.3057851239669422, | |
| "grad_norm": 1.1259503364562988, | |
| "learning_rate": 8.001609142396149e-05, | |
| "loss": 0.089, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.3884297520661155, | |
| "grad_norm": 1.8521549701690674, | |
| "learning_rate": 7.891165979136429e-05, | |
| "loss": 0.0741, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.4710743801652892, | |
| "grad_norm": 0.6957982778549194, | |
| "learning_rate": 7.778561562639818e-05, | |
| "loss": 0.076, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.553719008264463, | |
| "grad_norm": 0.4278518259525299, | |
| "learning_rate": 7.663880068862106e-05, | |
| "loss": 0.0752, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.9472731351852417, | |
| "learning_rate": 7.547207226450979e-05, | |
| "loss": 0.0899, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.71900826446281, | |
| "grad_norm": 0.472150593996048, | |
| "learning_rate": 7.428630252660704e-05, | |
| "loss": 0.0746, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.8016528925619832, | |
| "grad_norm": 0.614167332649231, | |
| "learning_rate": 7.308237788154003e-05, | |
| "loss": 0.0785, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.884297520661157, | |
| "grad_norm": 0.5245425701141357, | |
| "learning_rate": 7.186119830739883e-05, | |
| "loss": 0.0756, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.9669421487603307, | |
| "grad_norm": 0.4369754195213318, | |
| "learning_rate": 7.062367668096967e-05, | |
| "loss": 0.0673, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.049586776859504, | |
| "grad_norm": 0.7638271450996399, | |
| "learning_rate": 6.93707380953258e-05, | |
| "loss": 0.0582, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.132231404958677, | |
| "grad_norm": 0.7099161744117737, | |
| "learning_rate": 6.810331916828622e-05, | |
| "loss": 0.065, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.214876033057851, | |
| "grad_norm": 0.9365837574005127, | |
| "learning_rate": 6.682236734225944e-05, | |
| "loss": 0.0736, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.297520661157025, | |
| "grad_norm": 0.5708351135253906, | |
| "learning_rate": 6.552884017599517e-05, | |
| "loss": 0.0582, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.380165289256198, | |
| "grad_norm": 2.0133728981018066, | |
| "learning_rate": 6.422370462877396e-05, | |
| "loss": 0.0737, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.462809917355372, | |
| "grad_norm": 0.615628719329834, | |
| "learning_rate": 6.29079363375694e-05, | |
| "loss": 0.0619, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 1.595961570739746, | |
| "learning_rate": 6.15825188877235e-05, | |
| "loss": 0.0646, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.628099173553719, | |
| "grad_norm": 0.5787972807884216, | |
| "learning_rate": 6.0248443077680316e-05, | |
| "loss": 0.0727, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.710743801652892, | |
| "grad_norm": 0.37428799271583557, | |
| "learning_rate": 5.890670617832764e-05, | |
| "loss": 0.0694, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.793388429752066, | |
| "grad_norm": 0.35877710580825806, | |
| "learning_rate": 5.755831118750016e-05, | |
| "loss": 0.0615, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.87603305785124, | |
| "grad_norm": 0.7337440252304077, | |
| "learning_rate": 5.620426608020156e-05, | |
| "loss": 0.0693, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.958677685950414, | |
| "grad_norm": 0.4097314476966858, | |
| "learning_rate": 5.484558305510609e-05, | |
| "loss": 0.0584, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.041322314049586, | |
| "grad_norm": 0.9892732501029968, | |
| "learning_rate": 5.348327777790262e-05, | |
| "loss": 0.0654, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.12396694214876, | |
| "grad_norm": 0.9509891867637634, | |
| "learning_rate": 5.211836862204715e-05, | |
| "loss": 0.0594, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 5.206611570247934, | |
| "grad_norm": 0.6350495219230652, | |
| "learning_rate": 5.075187590749101e-05, | |
| "loss": 0.0568, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 5.289256198347108, | |
| "grad_norm": 0.4029739201068878, | |
| "learning_rate": 4.93848211379541e-05, | |
| "loss": 0.0572, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.371900826446281, | |
| "grad_norm": 0.8811182379722595, | |
| "learning_rate": 4.8018226237313165e-05, | |
| "loss": 0.0602, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.9591042995452881, | |
| "learning_rate": 4.665311278567593e-05, | |
| "loss": 0.0571, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.537190082644628, | |
| "grad_norm": 0.4754871726036072, | |
| "learning_rate": 4.5290501255712415e-05, | |
| "loss": 0.0489, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.619834710743802, | |
| "grad_norm": 1.2373918294906616, | |
| "learning_rate": 4.3931410249813806e-05, | |
| "loss": 0.0559, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.702479338842975, | |
| "grad_norm": 0.8961721062660217, | |
| "learning_rate": 4.2576855738649714e-05, | |
| "loss": 0.0532, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.785123966942149, | |
| "grad_norm": 0.4459110498428345, | |
| "learning_rate": 4.122785030169256e-05, | |
| "loss": 0.0655, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.867768595041323, | |
| "grad_norm": 0.38759222626686096, | |
| "learning_rate": 3.988540237027702e-05, | |
| "loss": 0.0542, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 5.950413223140496, | |
| "grad_norm": 1.9169152975082397, | |
| "learning_rate": 3.8550515473760514e-05, | |
| "loss": 0.0507, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 6.033057851239669, | |
| "grad_norm": 2.071638345718384, | |
| "learning_rate": 3.722418748934785e-05, | |
| "loss": 0.0559, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 6.115702479338843, | |
| "grad_norm": 0.7945961952209473, | |
| "learning_rate": 3.590740989614131e-05, | |
| "loss": 0.0542, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 6.198347107438017, | |
| "grad_norm": 0.488382488489151, | |
| "learning_rate": 3.460116703397336e-05, | |
| "loss": 0.0599, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.2809917355371905, | |
| "grad_norm": 0.5963483452796936, | |
| "learning_rate": 3.330643536757638e-05, | |
| "loss": 0.051, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.6972360610961914, | |
| "learning_rate": 3.2024182756639185e-05, | |
| "loss": 0.0574, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.446280991735537, | |
| "grad_norm": 0.6945924758911133, | |
| "learning_rate": 3.075536773229624e-05, | |
| "loss": 0.0533, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.528925619834711, | |
| "grad_norm": 0.3093964755535126, | |
| "learning_rate": 2.9500938780590275e-05, | |
| "loss": 0.0567, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.6115702479338845, | |
| "grad_norm": 1.643970251083374, | |
| "learning_rate": 2.826183363344391e-05, | |
| "loss": 0.0527, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.694214876033058, | |
| "grad_norm": 0.8777986168861389, | |
| "learning_rate": 2.7038978567670558e-05, | |
| "loss": 0.0543, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 6.776859504132231, | |
| "grad_norm": 0.5457648038864136, | |
| "learning_rate": 2.5833287712548198e-05, | |
| "loss": 0.0533, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 6.859504132231405, | |
| "grad_norm": 0.49185365438461304, | |
| "learning_rate": 2.4645662366474188e-05, | |
| "loss": 0.0551, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 6.9421487603305785, | |
| "grad_norm": 0.3884984254837036, | |
| "learning_rate": 2.3476990323211267e-05, | |
| "loss": 0.0513, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 7.024793388429752, | |
| "grad_norm": 0.5958797335624695, | |
| "learning_rate": 2.2328145208229095e-05, | |
| "loss": 0.0631, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.107438016528926, | |
| "grad_norm": 1.0943728685379028, | |
| "learning_rate": 2.119998582563692e-05, | |
| "loss": 0.0531, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 7.190082644628099, | |
| "grad_norm": 0.691371500492096, | |
| "learning_rate": 2.0093355516195888e-05, | |
| "loss": 0.0554, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 5.228278636932373, | |
| "learning_rate": 1.900908152689062e-05, | |
| "loss": 0.0509, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 7.355371900826446, | |
| "grad_norm": 0.3314078748226166, | |
| "learning_rate": 1.7947974392531612e-05, | |
| "loss": 0.0419, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 7.43801652892562, | |
| "grad_norm": 0.6459057927131653, | |
| "learning_rate": 1.6910827329850616e-05, | |
| "loss": 0.0443, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.520661157024794, | |
| "grad_norm": 0.29878008365631104, | |
| "learning_rate": 1.589841564454176e-05, | |
| "loss": 0.0424, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 7.6033057851239665, | |
| "grad_norm": 0.4006846249103546, | |
| "learning_rate": 1.4911496151692012e-05, | |
| "loss": 0.0479, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 7.68595041322314, | |
| "grad_norm": 0.5304248332977295, | |
| "learning_rate": 1.3950806610033957e-05, | |
| "loss": 0.0436, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 7.768595041322314, | |
| "grad_norm": 0.7876132726669312, | |
| "learning_rate": 1.3017065170443948e-05, | |
| "loss": 0.0448, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 7.851239669421488, | |
| "grad_norm": 0.9319537281990051, | |
| "learning_rate": 1.2110969839097797e-05, | |
| "loss": 0.045, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 7.933884297520661, | |
| "grad_norm": 1.810137391090393, | |
| "learning_rate": 1.1233197955685409e-05, | |
| "loss": 0.0432, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 8.016528925619834, | |
| "grad_norm": 1.6881096363067627, | |
| "learning_rate": 1.03844056870744e-05, | |
| "loss": 0.0492, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 8.099173553719009, | |
| "grad_norm": 1.5425840616226196, | |
| "learning_rate": 9.565227536801135e-06, | |
| "loss": 0.0454, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 1.0203477144241333, | |
| "learning_rate": 8.776275870755924e-06, | |
| "loss": 0.0551, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 8.264462809917354, | |
| "grad_norm": 1.097283959388733, | |
| "learning_rate": 8.018140459416962e-06, | |
| "loss": 0.039, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.347107438016529, | |
| "grad_norm": 0.6623409390449524, | |
| "learning_rate": 7.291388036975072e-06, | |
| "loss": 0.0438, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 8.429752066115702, | |
| "grad_norm": 0.38722532987594604, | |
| "learning_rate": 6.596561877679036e-06, | |
| "loss": 0.0541, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 8.512396694214877, | |
| "grad_norm": 0.1958349198102951, | |
| "learning_rate": 5.93418138971803e-06, | |
| "loss": 0.042, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 8.59504132231405, | |
| "grad_norm": 0.44341278076171875, | |
| "learning_rate": 5.304741726944873e-06, | |
| "loss": 0.0434, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 8.677685950413224, | |
| "grad_norm": 2.145524024963379, | |
| "learning_rate": 4.70871341873021e-06, | |
| "loss": 0.0357, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 8.760330578512397, | |
| "grad_norm": 1.4434798955917358, | |
| "learning_rate": 4.146542018224447e-06, | |
| "loss": 0.0405, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 8.84297520661157, | |
| "grad_norm": 0.6765860915184021, | |
| "learning_rate": 3.6186477692903954e-06, | |
| "loss": 0.0426, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 8.925619834710744, | |
| "grad_norm": 0.3515833020210266, | |
| "learning_rate": 3.1254252923553994e-06, | |
| "loss": 0.0464, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 9.008264462809917, | |
| "grad_norm": 0.4808438718318939, | |
| "learning_rate": 2.667243289418059e-06, | |
| "loss": 0.045, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.4166981875896454, | |
| "learning_rate": 2.244444268429857e-06, | |
| "loss": 0.0477, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.173553719008265, | |
| "grad_norm": 0.5427719950675964, | |
| "learning_rate": 1.8573442872578616e-06, | |
| "loss": 0.0494, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 9.256198347107437, | |
| "grad_norm": 0.3531731963157654, | |
| "learning_rate": 1.5062327174197644e-06, | |
| "loss": 0.0431, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 9.338842975206612, | |
| "grad_norm": 1.0387060642242432, | |
| "learning_rate": 1.191372027768034e-06, | |
| "loss": 0.0478, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 9.421487603305785, | |
| "grad_norm": 0.3243672251701355, | |
| "learning_rate": 9.129975882847364e-07, | |
| "loss": 0.0393, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 9.50413223140496, | |
| "grad_norm": 0.33455690741539, | |
| "learning_rate": 6.713174941338162e-07, | |
| "loss": 0.0419, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 9.586776859504132, | |
| "grad_norm": 0.4129684567451477, | |
| "learning_rate": 4.6651241010226e-07, | |
| "loss": 0.0463, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 9.669421487603305, | |
| "grad_norm": 2.450713872909546, | |
| "learning_rate": 2.9873543554652106e-07, | |
| "loss": 0.0448, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 9.75206611570248, | |
| "grad_norm": 0.22442331910133362, | |
| "learning_rate": 1.681119899450856e-07, | |
| "loss": 0.0462, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 9.834710743801653, | |
| "grad_norm": 0.9303910732269287, | |
| "learning_rate": 7.473971914280787e-08, | |
| "loss": 0.0449, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 9.917355371900827, | |
| "grad_norm": 1.7050765752792358, | |
| "learning_rate": 1.8688422357004966e-08, | |
| "loss": 0.0365, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.60956209897995, | |
| "learning_rate": 0.0, | |
| "loss": 0.0377, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 1210, | |
| "total_flos": 1.3059661832211168e+17, | |
| "train_loss": 0.09883986462738888, | |
| "train_runtime": 1198.7826, | |
| "train_samples_per_second": 49.142, | |
| "train_steps_per_second": 1.009 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3059661832211168e+17, | |
| "train_batch_size": 49, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |