| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 248, | |
| "global_step": 4960, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3e-05, | |
| "loss": 2.8428, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3e-05, | |
| "loss": 2.1589, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0527, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9823, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6398457059679767, | |
| "eval_loss": 1.6149228811264038, | |
| "eval_runtime": 17.4504, | |
| "eval_samples_per_second": 28.653, | |
| "eval_steps_per_second": 1.834, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 9.0, | |
| "eval_f1": 13.45666666666666, | |
| "eval_qa_bleu": 7.063432441196736, | |
| "eval_qa_exact_match": 0.09, | |
| "eval_recite_bleu": 15.344658855358656, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9153, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 3e-05, | |
| "loss": 1.49, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4406, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4161, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3868, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6700320232896652, | |
| "eval_loss": 1.1928881406784058, | |
| "eval_runtime": 15.4486, | |
| "eval_samples_per_second": 32.365, | |
| "eval_steps_per_second": 2.071, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 9.2, | |
| "eval_f1": 14.735098500533276, | |
| "eval_qa_bleu": 7.255412763224127, | |
| "eval_qa_exact_match": 0.088, | |
| "eval_recite_bleu": 17.979965007898493, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 3e-05, | |
| "loss": 1.309, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9401, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9429, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9115, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9102, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6992489082969432, | |
| "eval_loss": 0.8513368368148804, | |
| "eval_runtime": 15.7421, | |
| "eval_samples_per_second": 31.762, | |
| "eval_steps_per_second": 2.033, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 10.2, | |
| "eval_f1": 15.888450046685344, | |
| "eval_qa_bleu": 7.2379640900667646, | |
| "eval_qa_exact_match": 0.098, | |
| "eval_recite_bleu": 20.876580135938546, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8302, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5813, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5844, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5776, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5869, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.721787481804949, | |
| "eval_loss": 0.6180558204650879, | |
| "eval_runtime": 15.5183, | |
| "eval_samples_per_second": 32.22, | |
| "eval_steps_per_second": 2.062, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 13.6, | |
| "eval_f1": 19.457685255920552, | |
| "eval_qa_bleu": 7.372812651620524, | |
| "eval_qa_exact_match": 0.126, | |
| "eval_recite_bleu": 24.119469757278573, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5394, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3787, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3865, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3734, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "learning_rate": 3e-05, | |
| "loss": 0.364, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7352052401746725, | |
| "eval_loss": 0.49626532196998596, | |
| "eval_runtime": 15.762, | |
| "eval_samples_per_second": 31.722, | |
| "eval_steps_per_second": 2.03, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 11.6, | |
| "eval_f1": 17.050505050505052, | |
| "eval_qa_bleu": 9.3668737780184, | |
| "eval_qa_exact_match": 0.114, | |
| "eval_recite_bleu": 28.6471128712922, | |
| "eval_recite_exact_match": 0.004, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 3e-05, | |
| "loss": 0.345, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2508, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 3e-05, | |
| "loss": 0.26, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2618, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2706, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7419825327510917, | |
| "eval_loss": 0.441346138715744, | |
| "eval_runtime": 15.4217, | |
| "eval_samples_per_second": 32.422, | |
| "eval_steps_per_second": 2.075, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 17.6, | |
| "eval_f1": 23.964126984126988, | |
| "eval_qa_bleu": 13.1735935431327, | |
| "eval_qa_exact_match": 0.164, | |
| "eval_recite_bleu": 35.20769002447571, | |
| "eval_recite_exact_match": 0.008, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2459, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1802, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 6.45, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1853, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1905, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1906, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7457903930131005, | |
| "eval_loss": 0.4084840714931488, | |
| "eval_runtime": 15.7338, | |
| "eval_samples_per_second": 31.779, | |
| "eval_steps_per_second": 2.034, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 20.2, | |
| "eval_f1": 25.946536796536805, | |
| "eval_qa_bleu": 13.109935567750034, | |
| "eval_qa_exact_match": 0.198, | |
| "eval_recite_bleu": 41.370445416956926, | |
| "eval_recite_exact_match": 0.018, | |
| "step": 1736 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1735, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.26, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1402, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.46, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1423, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1468, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1474, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7476826783114993, | |
| "eval_loss": 0.4009522497653961, | |
| "eval_runtime": 15.0512, | |
| "eval_samples_per_second": 33.22, | |
| "eval_steps_per_second": 2.126, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 17.4, | |
| "eval_f1": 24.613044733044738, | |
| "eval_qa_bleu": 11.328428780013391, | |
| "eval_qa_exact_match": 0.162, | |
| "eval_recite_bleu": 45.62279446830165, | |
| "eval_recite_exact_match": 0.036, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1383, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.27, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1181, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1212, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 3e-05, | |
| "loss": 0.118, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 8.87, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1264, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7484366812227075, | |
| "eval_loss": 0.39788317680358887, | |
| "eval_runtime": 15.8038, | |
| "eval_samples_per_second": 31.638, | |
| "eval_steps_per_second": 2.025, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 16.4, | |
| "eval_f1": 23.748124098124105, | |
| "eval_qa_bleu": 10.821376307326117, | |
| "eval_qa_exact_match": 0.148, | |
| "eval_recite_bleu": 46.63269966666572, | |
| "eval_recite_exact_match": 0.04, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "learning_rate": 3e-05, | |
| "loss": 0.121, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.27, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1036, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1053, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1129, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1091, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7487336244541485, | |
| "eval_loss": 0.4059823751449585, | |
| "eval_runtime": 15.4716, | |
| "eval_samples_per_second": 32.317, | |
| "eval_steps_per_second": 2.068, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 16.6, | |
| "eval_f1": 24.02516483516484, | |
| "eval_qa_bleu": 9.045796919947849, | |
| "eval_qa_exact_match": 0.156, | |
| "eval_recite_bleu": 50.48708434203323, | |
| "eval_recite_exact_match": 0.042, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 10.08, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1062, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "learning_rate": 3e-05, | |
| "loss": 0.098, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 10.48, | |
| "learning_rate": 3e-05, | |
| "loss": 0.103, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.69, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1015, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 10.89, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1072, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.749018922852984, | |
| "eval_loss": 0.4050058424472809, | |
| "eval_runtime": 15.8782, | |
| "eval_samples_per_second": 31.49, | |
| "eval_steps_per_second": 2.015, | |
| "step": 2728 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_exact_match": 17.0, | |
| "eval_f1": 23.9401875901876, | |
| "eval_qa_bleu": 11.704401309760495, | |
| "eval_qa_exact_match": 0.154, | |
| "eval_recite_bleu": 49.91003032025176, | |
| "eval_recite_exact_match": 0.054, | |
| "step": 2728 | |
| }, | |
| { | |
| "epoch": 11.09, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1013, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.29, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0969, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0971, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 11.69, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0984, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0969, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7491644832605532, | |
| "eval_loss": 0.4080793559551239, | |
| "eval_runtime": 15.5323, | |
| "eval_samples_per_second": 32.191, | |
| "eval_steps_per_second": 2.06, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 18.2, | |
| "eval_f1": 25.5360361860362, | |
| "eval_qa_bleu": 14.03126374073827, | |
| "eval_qa_exact_match": 0.176, | |
| "eval_recite_bleu": 49.89862429102314, | |
| "eval_recite_exact_match": 0.064, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0977, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0917, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0927, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0963, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0935, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7494788937409025, | |
| "eval_loss": 0.4145064055919647, | |
| "eval_runtime": 15.6596, | |
| "eval_samples_per_second": 31.929, | |
| "eval_steps_per_second": 2.043, | |
| "step": 3224 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_exact_match": 19.4, | |
| "eval_f1": 26.85320346320347, | |
| "eval_qa_bleu": 12.354122306398153, | |
| "eval_qa_exact_match": 0.18, | |
| "eval_recite_bleu": 51.76883509669797, | |
| "eval_recite_exact_match": 0.068, | |
| "step": 3224 | |
| }, | |
| { | |
| "epoch": 13.1, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0897, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.31, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0905, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0906, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 13.71, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0929, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 13.91, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0932, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7494177583697235, | |
| "eval_loss": 0.4077925384044647, | |
| "eval_runtime": 15.5414, | |
| "eval_samples_per_second": 32.172, | |
| "eval_steps_per_second": 2.059, | |
| "step": 3472 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_exact_match": 20.0, | |
| "eval_f1": 27.0681240981241, | |
| "eval_qa_bleu": 11.488299016816185, | |
| "eval_qa_exact_match": 0.182, | |
| "eval_recite_bleu": 50.751064727490615, | |
| "eval_recite_exact_match": 0.072, | |
| "step": 3472 | |
| }, | |
| { | |
| "epoch": 14.11, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0918, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.31, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0855, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 14.52, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0918, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.72, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0951, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 14.92, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0929, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7494439592430859, | |
| "eval_loss": 0.41397154331207275, | |
| "eval_runtime": 15.6399, | |
| "eval_samples_per_second": 31.969, | |
| "eval_steps_per_second": 2.046, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_exact_match": 21.4, | |
| "eval_f1": 28.090981240981243, | |
| "eval_qa_bleu": 12.708765778160004, | |
| "eval_qa_exact_match": 0.204, | |
| "eval_recite_bleu": 51.72639724879611, | |
| "eval_recite_exact_match": 0.068, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 15.12, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0875, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.32, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0889, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.52, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0871, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 15.73, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0918, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 15.93, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0951, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7494730713245997, | |
| "eval_loss": 0.4144821763038635, | |
| "eval_runtime": 15.6836, | |
| "eval_samples_per_second": 31.88, | |
| "eval_steps_per_second": 2.04, | |
| "step": 3968 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_exact_match": 19.8, | |
| "eval_f1": 27.501433011433015, | |
| "eval_qa_bleu": 16.46590504145556, | |
| "eval_qa_exact_match": 0.186, | |
| "eval_recite_bleu": 52.1066910776491, | |
| "eval_recite_exact_match": 0.074, | |
| "step": 3968 | |
| }, | |
| { | |
| "epoch": 16.13, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0818, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0894, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 16.53, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0897, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.73, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0914, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 16.94, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0926, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7494847161572052, | |
| "eval_loss": 0.4133778512477875, | |
| "eval_runtime": 15.6296, | |
| "eval_samples_per_second": 31.991, | |
| "eval_steps_per_second": 2.047, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_exact_match": 19.2, | |
| "eval_f1": 26.86869424692955, | |
| "eval_qa_bleu": 15.6053335659073, | |
| "eval_qa_exact_match": 0.178, | |
| "eval_recite_bleu": 52.354652374608186, | |
| "eval_recite_exact_match": 0.068, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 17.14, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0884, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.34, | |
| "learning_rate": 3e-05, | |
| "loss": 0.085, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0908, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 17.74, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0898, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 17.94, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0946, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7493420669577875, | |
| "eval_loss": 0.42565715312957764, | |
| "eval_runtime": 15.8896, | |
| "eval_samples_per_second": 31.467, | |
| "eval_steps_per_second": 2.014, | |
| "step": 4464 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_exact_match": 21.2, | |
| "eval_f1": 29.34194877018406, | |
| "eval_qa_bleu": 13.32335745765551, | |
| "eval_qa_exact_match": 0.196, | |
| "eval_recite_bleu": 53.51642016865907, | |
| "eval_recite_exact_match": 0.074, | |
| "step": 4464 | |
| }, | |
| { | |
| "epoch": 18.15, | |
| "learning_rate": 3e-05, | |
| "loss": 0.09, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.35, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0883, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 18.55, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0866, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0901, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 18.95, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0897, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7495778748180495, | |
| "eval_loss": 0.41643163561820984, | |
| "eval_runtime": 15.7007, | |
| "eval_samples_per_second": 31.846, | |
| "eval_steps_per_second": 2.038, | |
| "step": 4712 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_exact_match": 20.2, | |
| "eval_f1": 27.71383838383839, | |
| "eval_qa_bleu": 14.36873555788707, | |
| "eval_qa_exact_match": 0.192, | |
| "eval_recite_bleu": 50.76206757901549, | |
| "eval_recite_exact_match": 0.072, | |
| "step": 4712 | |
| }, | |
| { | |
| "epoch": 19.15, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0858, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.35, | |
| "learning_rate": 3e-05, | |
| "loss": 0.086, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.56, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0862, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 19.76, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0876, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 19.96, | |
| "learning_rate": 3e-05, | |
| "loss": 0.092, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7496011644832605, | |
| "eval_loss": 0.4180622398853302, | |
| "eval_runtime": 15.5975, | |
| "eval_samples_per_second": 32.056, | |
| "eval_steps_per_second": 2.052, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_exact_match": 21.4, | |
| "eval_f1": 29.01558441558442, | |
| "eval_qa_bleu": 14.570041032641159, | |
| "eval_qa_exact_match": 0.2, | |
| "eval_recite_bleu": 53.657481681296964, | |
| "eval_recite_exact_match": 0.072, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 4960, | |
| "total_flos": 2.30297745517824e+17, | |
| "train_loss": 0.3613291465947705, | |
| "train_runtime": 9979.1636, | |
| "train_samples_per_second": 7.935, | |
| "train_steps_per_second": 0.497 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4960, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 2.30297745517824e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |