|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 38.0, |
|
"eval_steps": 500, |
|
"global_step": 9500, |
|
"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 |
|
}, |
|
{ |
|
"epoch": 21.2, |
|
"grad_norm": 1.4770017862319946, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3518, |
|
"step": 5300 |
|
}, |
|
{ |
|
"epoch": 21.6, |
|
"grad_norm": 0.7377692461013794, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3498, |
|
"step": 5400 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"grad_norm": 0.7583816051483154, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3639, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_accuracy": 0.49098412698412697, |
|
"eval_loss": 3.951629400253296, |
|
"eval_runtime": 5.1413, |
|
"eval_samples_per_second": 97.251, |
|
"eval_steps_per_second": 12.254, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_exact_match": 16.2, |
|
"eval_f1": 22.613965478965486, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 22.4, |
|
"grad_norm": 2.125830888748169, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3386, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 22.8, |
|
"grad_norm": 0.958363950252533, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3533, |
|
"step": 5700 |
|
}, |
|
{ |
|
"epoch": 23.0, |
|
"eval_accuracy": 0.4915555555555556, |
|
"eval_loss": 4.020674228668213, |
|
"eval_runtime": 5.154, |
|
"eval_samples_per_second": 97.011, |
|
"eval_steps_per_second": 12.223, |
|
"step": 5750 |
|
}, |
|
{ |
|
"epoch": 23.0, |
|
"eval_exact_match": 16.6, |
|
"eval_f1": 21.791541236541246, |
|
"step": 5750 |
|
}, |
|
{ |
|
"epoch": 23.2, |
|
"grad_norm": 1.526780605316162, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3437, |
|
"step": 5800 |
|
}, |
|
{ |
|
"epoch": 23.6, |
|
"grad_norm": 1.0150421857833862, |
|
"learning_rate": 0.0001, |
|
"loss": 0.341, |
|
"step": 5900 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"grad_norm": 0.9590368866920471, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3587, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.49165079365079367, |
|
"eval_loss": 3.990480661392212, |
|
"eval_runtime": 5.2849, |
|
"eval_samples_per_second": 94.608, |
|
"eval_steps_per_second": 11.921, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_exact_match": 14.8, |
|
"eval_f1": 20.68870512528408, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 24.4, |
|
"grad_norm": 0.7218676209449768, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3355, |
|
"step": 6100 |
|
}, |
|
{ |
|
"epoch": 24.8, |
|
"grad_norm": 1.0547350645065308, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3479, |
|
"step": 6200 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"eval_accuracy": 0.4914920634920635, |
|
"eval_loss": 4.061710357666016, |
|
"eval_runtime": 5.1785, |
|
"eval_samples_per_second": 96.553, |
|
"eval_steps_per_second": 12.166, |
|
"step": 6250 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"eval_exact_match": 15.2, |
|
"eval_f1": 21.924090149616465, |
|
"step": 6250 |
|
}, |
|
{ |
|
"epoch": 25.2, |
|
"grad_norm": 0.8290365934371948, |
|
"learning_rate": 0.0001, |
|
"loss": 0.339, |
|
"step": 6300 |
|
}, |
|
{ |
|
"epoch": 25.6, |
|
"grad_norm": 1.2108442783355713, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3409, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"grad_norm": 1.596282720565796, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3511, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"eval_accuracy": 0.4903174603174603, |
|
"eval_loss": 4.0105509757995605, |
|
"eval_runtime": 5.1332, |
|
"eval_samples_per_second": 97.405, |
|
"eval_steps_per_second": 12.273, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"eval_exact_match": 15.2, |
|
"eval_f1": 22.125512265512278, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 26.4, |
|
"grad_norm": 0.607507050037384, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3289, |
|
"step": 6600 |
|
}, |
|
{ |
|
"epoch": 26.8, |
|
"grad_norm": 0.8661500215530396, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3442, |
|
"step": 6700 |
|
}, |
|
{ |
|
"epoch": 27.0, |
|
"eval_accuracy": 0.491015873015873, |
|
"eval_loss": 4.040103435516357, |
|
"eval_runtime": 5.3436, |
|
"eval_samples_per_second": 93.57, |
|
"eval_steps_per_second": 11.79, |
|
"step": 6750 |
|
}, |
|
{ |
|
"epoch": 27.0, |
|
"eval_exact_match": 15.8, |
|
"eval_f1": 23.22610556110557, |
|
"step": 6750 |
|
}, |
|
{ |
|
"epoch": 27.2, |
|
"grad_norm": 0.6577064394950867, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3417, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 27.6, |
|
"grad_norm": 0.7707544565200806, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3367, |
|
"step": 6900 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"grad_norm": 0.9173874258995056, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3496, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.48968253968253966, |
|
"eval_loss": 4.015742778778076, |
|
"eval_runtime": 5.5561, |
|
"eval_samples_per_second": 89.991, |
|
"eval_steps_per_second": 11.339, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_exact_match": 14.8, |
|
"eval_f1": 21.22861116361117, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 28.4, |
|
"grad_norm": 0.8520764112472534, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3324, |
|
"step": 7100 |
|
}, |
|
{ |
|
"epoch": 28.8, |
|
"grad_norm": 0.6646843552589417, |
|
"learning_rate": 0.0001, |
|
"loss": 0.34, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 29.0, |
|
"eval_accuracy": 0.4902222222222222, |
|
"eval_loss": 4.0503458976745605, |
|
"eval_runtime": 5.543, |
|
"eval_samples_per_second": 90.204, |
|
"eval_steps_per_second": 11.366, |
|
"step": 7250 |
|
}, |
|
{ |
|
"epoch": 29.0, |
|
"eval_exact_match": 15.4, |
|
"eval_f1": 21.908746993747, |
|
"step": 7250 |
|
}, |
|
{ |
|
"epoch": 29.2, |
|
"grad_norm": 0.5370341539382935, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3325, |
|
"step": 7300 |
|
}, |
|
{ |
|
"epoch": 29.6, |
|
"grad_norm": 0.9284526705741882, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3353, |
|
"step": 7400 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"grad_norm": 1.1908068656921387, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3448, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"eval_accuracy": 0.49082539682539683, |
|
"eval_loss": 4.078609466552734, |
|
"eval_runtime": 5.4942, |
|
"eval_samples_per_second": 91.006, |
|
"eval_steps_per_second": 11.467, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"eval_exact_match": 14.6, |
|
"eval_f1": 21.577283272283278, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 30.4, |
|
"grad_norm": 0.8314960598945618, |
|
"learning_rate": 0.0001, |
|
"loss": 0.325, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 30.8, |
|
"grad_norm": 0.8316343426704407, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3406, |
|
"step": 7700 |
|
}, |
|
{ |
|
"epoch": 31.0, |
|
"eval_accuracy": 0.49053968253968255, |
|
"eval_loss": 4.123876571655273, |
|
"eval_runtime": 5.1277, |
|
"eval_samples_per_second": 97.51, |
|
"eval_steps_per_second": 12.286, |
|
"step": 7750 |
|
}, |
|
{ |
|
"epoch": 31.0, |
|
"eval_exact_match": 15.0, |
|
"eval_f1": 21.135220890220893, |
|
"step": 7750 |
|
}, |
|
{ |
|
"epoch": 31.2, |
|
"grad_norm": 0.539159893989563, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3354, |
|
"step": 7800 |
|
}, |
|
{ |
|
"epoch": 31.6, |
|
"grad_norm": 0.7684648633003235, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3354, |
|
"step": 7900 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"grad_norm": 0.8351700901985168, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3375, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.4914603174603175, |
|
"eval_loss": 4.1210455894470215, |
|
"eval_runtime": 5.5561, |
|
"eval_samples_per_second": 89.991, |
|
"eval_steps_per_second": 11.339, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_exact_match": 14.4, |
|
"eval_f1": 20.92217171717173, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 32.4, |
|
"grad_norm": 0.9775452017784119, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3233, |
|
"step": 8100 |
|
}, |
|
{ |
|
"epoch": 32.8, |
|
"grad_norm": 0.8110117316246033, |
|
"learning_rate": 0.0001, |
|
"loss": 0.339, |
|
"step": 8200 |
|
}, |
|
{ |
|
"epoch": 33.0, |
|
"eval_accuracy": 0.48984126984126986, |
|
"eval_loss": 4.103898525238037, |
|
"eval_runtime": 5.2522, |
|
"eval_samples_per_second": 95.197, |
|
"eval_steps_per_second": 11.995, |
|
"step": 8250 |
|
}, |
|
{ |
|
"epoch": 33.0, |
|
"eval_exact_match": 15.6, |
|
"eval_f1": 22.6055772005772, |
|
"step": 8250 |
|
}, |
|
{ |
|
"epoch": 33.2, |
|
"grad_norm": 0.617221474647522, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3304, |
|
"step": 8300 |
|
}, |
|
{ |
|
"epoch": 33.6, |
|
"grad_norm": 0.7995471358299255, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3294, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"grad_norm": 0.7901666760444641, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3418, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"eval_accuracy": 0.4901904761904762, |
|
"eval_loss": 4.087856769561768, |
|
"eval_runtime": 5.2835, |
|
"eval_samples_per_second": 94.634, |
|
"eval_steps_per_second": 11.924, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"eval_exact_match": 15.4, |
|
"eval_f1": 22.837828282828294, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 34.4, |
|
"grad_norm": 0.7124672532081604, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3227, |
|
"step": 8600 |
|
}, |
|
{ |
|
"epoch": 34.8, |
|
"grad_norm": 0.9989972114562988, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3364, |
|
"step": 8700 |
|
}, |
|
{ |
|
"epoch": 35.0, |
|
"eval_accuracy": 0.4906984126984127, |
|
"eval_loss": 4.07823371887207, |
|
"eval_runtime": 5.4413, |
|
"eval_samples_per_second": 91.889, |
|
"eval_steps_per_second": 11.578, |
|
"step": 8750 |
|
}, |
|
{ |
|
"epoch": 35.0, |
|
"eval_exact_match": 16.0, |
|
"eval_f1": 22.744945200997837, |
|
"step": 8750 |
|
}, |
|
{ |
|
"epoch": 35.2, |
|
"grad_norm": 1.1887849569320679, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3279, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 35.6, |
|
"grad_norm": 1.1754401922225952, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3301, |
|
"step": 8900 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"grad_norm": 1.0393660068511963, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3421, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.49098412698412697, |
|
"eval_loss": 4.051205635070801, |
|
"eval_runtime": 5.1509, |
|
"eval_samples_per_second": 97.07, |
|
"eval_steps_per_second": 12.231, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_exact_match": 15.6, |
|
"eval_f1": 22.011782106782118, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 36.4, |
|
"grad_norm": 0.5955029129981995, |
|
"learning_rate": 0.0001, |
|
"loss": 0.322, |
|
"step": 9100 |
|
}, |
|
{ |
|
"epoch": 36.8, |
|
"grad_norm": 0.5837319493293762, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3337, |
|
"step": 9200 |
|
}, |
|
{ |
|
"epoch": 37.0, |
|
"eval_accuracy": 0.4894603174603175, |
|
"eval_loss": 4.1726765632629395, |
|
"eval_runtime": 5.8414, |
|
"eval_samples_per_second": 85.595, |
|
"eval_steps_per_second": 10.785, |
|
"step": 9250 |
|
}, |
|
{ |
|
"epoch": 37.0, |
|
"eval_exact_match": 15.0, |
|
"eval_f1": 21.872969807969813, |
|
"step": 9250 |
|
}, |
|
{ |
|
"epoch": 37.2, |
|
"grad_norm": 0.8081912994384766, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3272, |
|
"step": 9300 |
|
}, |
|
{ |
|
"epoch": 37.6, |
|
"grad_norm": 0.8411530256271362, |
|
"learning_rate": 0.0001, |
|
"loss": 0.331, |
|
"step": 9400 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"grad_norm": 1.546312689781189, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3375, |
|
"step": 9500 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"eval_accuracy": 0.4889206349206349, |
|
"eval_loss": 4.161491870880127, |
|
"eval_runtime": 5.1783, |
|
"eval_samples_per_second": 96.557, |
|
"eval_steps_per_second": 12.166, |
|
"step": 9500 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"eval_exact_match": 15.6, |
|
"eval_f1": 22.188371073371083, |
|
"step": 9500 |
|
} |
|
], |
|
"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": 2.9063767753805005e+17, |
|
"train_batch_size": 2, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|