| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 21.0, | |
| "eval_steps": 500, | |
| "global_step": 5250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.5897082090377808, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3566, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5390245914459229, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2503, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5155873015873016, | |
| "eval_loss": 2.3237216472625732, | |
| "eval_runtime": 6.3164, | |
| "eval_samples_per_second": 79.159, | |
| "eval_steps_per_second": 9.974, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 17.2, | |
| "eval_f1": 25.586581359816655, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.7468857169151306, | |
| "learning_rate": 0.0001, | |
| "loss": 2.1798, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.9109599590301514, | |
| "learning_rate": 0.0001, | |
| "loss": 2.1094, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8458774089813232, | |
| "learning_rate": 0.0001, | |
| "loss": 2.087, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5163809523809524, | |
| "eval_loss": 2.3309242725372314, | |
| "eval_runtime": 5.1896, | |
| "eval_samples_per_second": 96.347, | |
| "eval_steps_per_second": 12.14, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 18.6, | |
| "eval_f1": 26.043196248196264, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.171212077140808, | |
| "learning_rate": 0.0001, | |
| "loss": 1.8413, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.1294909715652466, | |
| "learning_rate": 0.0001, | |
| "loss": 1.849, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5144761904761905, | |
| "eval_loss": 2.401942253112793, | |
| "eval_runtime": 5.5775, | |
| "eval_samples_per_second": 89.646, | |
| "eval_steps_per_second": 11.295, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 18.4, | |
| "eval_f1": 26.966370002546494, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.5013824701309204, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7322, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.4516321420669556, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5908, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.6037466526031494, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6193, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5103809523809524, | |
| "eval_loss": 2.5039021968841553, | |
| "eval_runtime": 5.1568, | |
| "eval_samples_per_second": 96.96, | |
| "eval_steps_per_second": 12.217, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 18.4, | |
| "eval_f1": 26.892813852813866, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.7909170389175415, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3298, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.9236161708831787, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3666, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.505015873015873, | |
| "eval_loss": 2.6543924808502197, | |
| "eval_runtime": 5.5115, | |
| "eval_samples_per_second": 90.719, | |
| "eval_steps_per_second": 11.431, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 18.4, | |
| "eval_f1": 25.875286609941792, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 1.8524874448776245, | |
| "learning_rate": 0.0001, | |
| "loss": 1.2381, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 2.1092071533203125, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1141, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.065767288208008, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1435, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5010793650793651, | |
| "eval_loss": 2.843566656112671, | |
| "eval_runtime": 5.1394, | |
| "eval_samples_per_second": 97.287, | |
| "eval_steps_per_second": 12.258, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 17.4, | |
| "eval_f1": 24.31514874014875, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 2.329406261444092, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8887, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 2.489598274230957, | |
| "learning_rate": 0.0001, | |
| "loss": 0.9171, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.49707936507936507, | |
| "eval_loss": 3.0319864749908447, | |
| "eval_runtime": 5.3015, | |
| "eval_samples_per_second": 94.313, | |
| "eval_steps_per_second": 11.883, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 16.0, | |
| "eval_f1": 22.824985569985575, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 2.764033794403076, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8093, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 2.5643441677093506, | |
| "learning_rate": 0.0001, | |
| "loss": 0.724, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.074833631515503, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7531, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.49304761904761907, | |
| "eval_loss": 3.2584712505340576, | |
| "eval_runtime": 5.2945, | |
| "eval_samples_per_second": 94.438, | |
| "eval_steps_per_second": 11.899, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 21.895264180264185, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 2.649679660797119, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5777, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 2.1969234943389893, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6101, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.4925079365079365, | |
| "eval_loss": 3.3418095111846924, | |
| "eval_runtime": 5.1381, | |
| "eval_samples_per_second": 97.312, | |
| "eval_steps_per_second": 12.261, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 21.76202020202021, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 1.8318747282028198, | |
| "learning_rate": 0.0001, | |
| "loss": 0.555, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 1.9915804862976074, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5102, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.1287789344787598, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5392, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.4915555555555556, | |
| "eval_loss": 3.537275791168213, | |
| "eval_runtime": 5.1264, | |
| "eval_samples_per_second": 97.534, | |
| "eval_steps_per_second": 12.289, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 15.4, | |
| "eval_f1": 22.23013114336644, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 1.6748021841049194, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4426, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 1.8744910955429077, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4718, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.48926984126984124, | |
| "eval_loss": 3.631282091140747, | |
| "eval_runtime": 5.3142, | |
| "eval_samples_per_second": 94.087, | |
| "eval_steps_per_second": 11.855, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_exact_match": 15.6, | |
| "eval_f1": 21.87386002886003, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 2.1398115158081055, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4456, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 1.8851490020751953, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4328, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.6958885192871094, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4446, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.49057142857142855, | |
| "eval_loss": 3.6736366748809814, | |
| "eval_runtime": 5.8652, | |
| "eval_samples_per_second": 85.248, | |
| "eval_steps_per_second": 10.741, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 15.2, | |
| "eval_f1": 21.81304966962862, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 1.8530863523483276, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3969, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 1.9440696239471436, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4204, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.4906031746031746, | |
| "eval_loss": 3.7342090606689453, | |
| "eval_runtime": 5.4947, | |
| "eval_samples_per_second": 90.996, | |
| "eval_steps_per_second": 11.466, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_exact_match": 15.8, | |
| "eval_f1": 21.908780663780668, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "grad_norm": 1.7972825765609741, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4036, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "grad_norm": 1.6307235956192017, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3929, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 2.7257182598114014, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4131, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.4897142857142857, | |
| "eval_loss": 3.7777669429779053, | |
| "eval_runtime": 5.8394, | |
| "eval_samples_per_second": 85.626, | |
| "eval_steps_per_second": 10.789, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 21.873367129946075, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 1.1628468036651611, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3756, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "grad_norm": 2.109792470932007, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3924, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.48965079365079367, | |
| "eval_loss": 3.820950746536255, | |
| "eval_runtime": 5.3214, | |
| "eval_samples_per_second": 93.96, | |
| "eval_steps_per_second": 11.839, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_exact_match": 14.6, | |
| "eval_f1": 22.25867243867245, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "grad_norm": 1.2581037282943726, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3792, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "grad_norm": 1.0244629383087158, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3767, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.3548619747161865, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3913, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.49038095238095236, | |
| "eval_loss": 3.8832576274871826, | |
| "eval_runtime": 5.4233, | |
| "eval_samples_per_second": 92.195, | |
| "eval_steps_per_second": 11.617, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 21.823196411431713, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "grad_norm": 1.5730881690979004, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3606, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "grad_norm": 1.619290828704834, | |
| "learning_rate": 0.0001, | |
| "loss": 0.376, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.4898095238095238, | |
| "eval_loss": 3.8935556411743164, | |
| "eval_runtime": 5.3789, | |
| "eval_samples_per_second": 92.957, | |
| "eval_steps_per_second": 11.713, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_exact_match": 15.2, | |
| "eval_f1": 21.284107517869135, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "grad_norm": 2.23779559135437, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3697, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "grad_norm": 1.2932429313659668, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3618, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.6087582111358643, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3785, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.49, | |
| "eval_loss": 3.882444381713867, | |
| "eval_runtime": 5.3456, | |
| "eval_samples_per_second": 93.536, | |
| "eval_steps_per_second": 11.785, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_exact_match": 15.4, | |
| "eval_f1": 21.950805305805307, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "grad_norm": 1.20901620388031, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3532, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "grad_norm": 1.3064277172088623, | |
| "learning_rate": 0.0001, | |
| "loss": 0.367, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.4900952380952381, | |
| "eval_loss": 3.972047805786133, | |
| "eval_runtime": 5.1196, | |
| "eval_samples_per_second": 97.664, | |
| "eval_steps_per_second": 12.306, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_exact_match": 16.4, | |
| "eval_f1": 23.046717171717184, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "grad_norm": 1.4567475318908691, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3578, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "grad_norm": 1.274679183959961, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3588, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.5081045627593994, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3676, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.4908888888888889, | |
| "eval_loss": 3.937361717224121, | |
| "eval_runtime": 5.3304, | |
| "eval_samples_per_second": 93.801, | |
| "eval_steps_per_second": 11.819, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_exact_match": 16.2, | |
| "eval_f1": 22.572724497724508, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.4, | |
| "grad_norm": 1.343906044960022, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3445, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "grad_norm": 1.7303237915039062, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3602, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.49044444444444446, | |
| "eval_loss": 3.9380335807800293, | |
| "eval_runtime": 5.2455, | |
| "eval_samples_per_second": 95.32, | |
| "eval_steps_per_second": 12.01, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_exact_match": 16.4, | |
| "eval_f1": 23.159069264069274, | |
| "step": 5250 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 12500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.606497816417403e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |