| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "global_step": 24100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 92.256, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 52.0764, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 6e-06, | |
| "loss": 41.3627, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 37.237, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 1e-05, | |
| "loss": 34.3031, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 1.2e-05, | |
| "loss": 31.7147, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 29.5475, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 27.914, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.8e-05, | |
| "loss": 26.1499, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2e-05, | |
| "loss": 24.1629, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 22.1829, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.4e-05, | |
| "loss": 20.2111, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 18.6642, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 16.7946, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 3e-05, | |
| "loss": 15.0783, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 13.366, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 11.706, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 3.6e-05, | |
| "loss": 10.2113, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 3.8e-05, | |
| "loss": 8.5921, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4e-05, | |
| "loss": 7.2445, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_bleu": 1.523, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 204.48, | |
| "eval_loss": 7.561939239501953, | |
| "eval_rm": NaN, | |
| "eval_runtime": 100.5427, | |
| "eval_samples_per_second": 0.497, | |
| "eval_steps_per_second": 0.07, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 4.2e-05, | |
| "loss": 6.8461, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 6.223, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 6.1167, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 5.7623, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 5e-05, | |
| "loss": 5.5056, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 4.976851851851852e-05, | |
| "loss": 5.2979, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 4.9537037037037035e-05, | |
| "loss": 4.9765, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 4.930555555555556e-05, | |
| "loss": 4.7721, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 4.9074074074074075e-05, | |
| "loss": 4.6957, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 4.8842592592592595e-05, | |
| "loss": 4.7368, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 4.6074, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 4.837962962962963e-05, | |
| "loss": 4.3678, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 4.19, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 4.791666666666667e-05, | |
| "loss": 4.0959, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 4.768518518518519e-05, | |
| "loss": 4.0323, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 4.745370370370371e-05, | |
| "loss": 3.7777, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 3.8014, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 4.699074074074074e-05, | |
| "loss": 3.6346, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.675925925925926e-05, | |
| "loss": 3.4876, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 4.652777777777778e-05, | |
| "loss": 3.4458, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_bleu": 16.0372, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 76.36, | |
| "eval_loss": 3.4201648235321045, | |
| "eval_rm": NaN, | |
| "eval_runtime": 91.7678, | |
| "eval_samples_per_second": 0.545, | |
| "eval_steps_per_second": 0.076, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 3.4425, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 4.6064814814814814e-05, | |
| "loss": 3.3346, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 4.5833333333333334e-05, | |
| "loss": 3.2582, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 4.5601851851851854e-05, | |
| "loss": 3.1174, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 4.5370370370370374e-05, | |
| "loss": 2.9873, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 4.5138888888888894e-05, | |
| "loss": 3.1041, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 4.490740740740741e-05, | |
| "loss": 2.9365, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 4.467592592592593e-05, | |
| "loss": 2.9494, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 2.8309, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 4.4212962962962966e-05, | |
| "loss": 2.6814, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 4.3981481481481486e-05, | |
| "loss": 2.6331, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 4.375e-05, | |
| "loss": 2.599, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 4.351851851851852e-05, | |
| "loss": 2.5576, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 4.328703703703704e-05, | |
| "loss": 2.4838, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 4.305555555555556e-05, | |
| "loss": 2.4532, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.282407407407408e-05, | |
| "loss": 2.3583, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 2.4245, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 4.236111111111111e-05, | |
| "loss": 2.3121, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 4.212962962962963e-05, | |
| "loss": 2.3254, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 4.1898148148148145e-05, | |
| "loss": 2.2571, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "eval_bleu": 34.5826, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 48.98, | |
| "eval_loss": 2.197143316268921, | |
| "eval_rm": 0.8125, | |
| "eval_runtime": 79.1278, | |
| "eval_samples_per_second": 0.632, | |
| "eval_steps_per_second": 0.088, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 2.2655, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 4.1435185185185185e-05, | |
| "loss": 2.1366, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 4.1203703703703705e-05, | |
| "loss": 2.1777, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 4.0972222222222225e-05, | |
| "loss": 2.1166, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 2.0804, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 4.0509259259259265e-05, | |
| "loss": 2.1024, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.027777777777778e-05, | |
| "loss": 2.0787, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 4.00462962962963e-05, | |
| "loss": 1.9875, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 3.981481481481482e-05, | |
| "loss": 1.9653, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 3.958333333333333e-05, | |
| "loss": 1.9254, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 3.935185185185186e-05, | |
| "loss": 1.9232, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 3.912037037037037e-05, | |
| "loss": 1.8927, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 1.7833, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 3.865740740740741e-05, | |
| "loss": 1.6375, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 3.8425925925925924e-05, | |
| "loss": 1.7233, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 1.6997, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 3.7962962962962964e-05, | |
| "loss": 1.746, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "learning_rate": 3.7731481481481484e-05, | |
| "loss": 1.6235, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.564, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 3.726851851851852e-05, | |
| "loss": 1.5696, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "eval_bleu": 53.8822, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 48.7, | |
| "eval_loss": 1.680190086364746, | |
| "eval_rm": 1.0, | |
| "eval_runtime": 52.3216, | |
| "eval_samples_per_second": 0.956, | |
| "eval_steps_per_second": 0.134, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.5516, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 3.6805555555555556e-05, | |
| "loss": 1.5438, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 3.6574074074074076e-05, | |
| "loss": 1.5262, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 3.6342592592592596e-05, | |
| "loss": 1.5738, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 1.5799, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 3.587962962962963e-05, | |
| "loss": 1.5116, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 3.564814814814815e-05, | |
| "loss": 1.5699, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 3.541666666666667e-05, | |
| "loss": 1.4994, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 1.3772, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 3.49537037037037e-05, | |
| "loss": 1.4322, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 1.4604, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 3.449074074074074e-05, | |
| "loss": 1.3924, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 3.425925925925926e-05, | |
| "loss": 1.348, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 3.402777777777778e-05, | |
| "loss": 1.3014, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 3.3796296296296295e-05, | |
| "loss": 1.3731, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 3.3564814814814815e-05, | |
| "loss": 1.39, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.2038, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 3.3101851851851855e-05, | |
| "loss": 1.1698, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 3.2870370370370375e-05, | |
| "loss": 1.1514, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 3.263888888888889e-05, | |
| "loss": 1.1359, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "eval_bleu": 64.5591, | |
| "eval_em": 0.02, | |
| "eval_gen_len": 45.2, | |
| "eval_loss": 1.346365213394165, | |
| "eval_rm": 0.973, | |
| "eval_runtime": 47.5326, | |
| "eval_samples_per_second": 1.052, | |
| "eval_steps_per_second": 0.147, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "learning_rate": 3.240740740740741e-05, | |
| "loss": 1.1396, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 3.217592592592593e-05, | |
| "loss": 1.1749, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 3.194444444444444e-05, | |
| "loss": 1.1098, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "learning_rate": 3.171296296296297e-05, | |
| "loss": 1.0741, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 1.1388, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3.125e-05, | |
| "loss": 1.0168, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 3.101851851851852e-05, | |
| "loss": 1.0862, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 3.0787037037037034e-05, | |
| "loss": 1.0689, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 3.055555555555556e-05, | |
| "loss": 1.0804, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "learning_rate": 3.0324074074074077e-05, | |
| "loss": 1.0312, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 3.0092592592592593e-05, | |
| "loss": 1.0184, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "learning_rate": 2.9861111111111113e-05, | |
| "loss": 0.992, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 1.0191, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "learning_rate": 2.9398148148148146e-05, | |
| "loss": 1.0842, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "learning_rate": 2.916666666666667e-05, | |
| "loss": 1.0508, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 2.8935185185185186e-05, | |
| "loss": 0.9815, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "learning_rate": 2.8703703703703706e-05, | |
| "loss": 0.9645, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 2.8472222222222223e-05, | |
| "loss": 0.9826, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 2.824074074074074e-05, | |
| "loss": 1.0036, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 2.8009259259259263e-05, | |
| "loss": 0.9994, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "eval_bleu": 68.0869, | |
| "eval_em": 0.02, | |
| "eval_gen_len": 47.76, | |
| "eval_loss": 1.0576136112213135, | |
| "eval_rm": 0.8889, | |
| "eval_runtime": 48.2259, | |
| "eval_samples_per_second": 1.037, | |
| "eval_steps_per_second": 0.145, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.8445, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "learning_rate": 2.75462962962963e-05, | |
| "loss": 0.833, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 2.7314814814814816e-05, | |
| "loss": 0.8034, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "learning_rate": 2.7083333333333332e-05, | |
| "loss": 0.8154, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 2.6851851851851855e-05, | |
| "loss": 0.8026, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "learning_rate": 2.6620370370370372e-05, | |
| "loss": 0.8147, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 2.6388888888888892e-05, | |
| "loss": 0.8466, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "learning_rate": 2.615740740740741e-05, | |
| "loss": 0.7881, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.7856, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "learning_rate": 2.5694444444444445e-05, | |
| "loss": 0.8058, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 2.5462962962962965e-05, | |
| "loss": 0.7704, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "learning_rate": 2.5231481481481485e-05, | |
| "loss": 0.7572, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.7757, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 2.4768518518518518e-05, | |
| "loss": 0.7598, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 2.4537037037037038e-05, | |
| "loss": 0.7268, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "learning_rate": 2.4305555555555558e-05, | |
| "loss": 0.7455, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.7525, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "learning_rate": 2.3842592592592594e-05, | |
| "loss": 0.7205, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 2.361111111111111e-05, | |
| "loss": 0.7741, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "learning_rate": 2.337962962962963e-05, | |
| "loss": 0.7275, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "eval_bleu": 74.1032, | |
| "eval_em": 0.02, | |
| "eval_gen_len": 46.52, | |
| "eval_loss": 0.952226996421814, | |
| "eval_rm": 0.9556, | |
| "eval_runtime": 45.6891, | |
| "eval_samples_per_second": 1.094, | |
| "eval_steps_per_second": 0.153, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "learning_rate": 2.314814814814815e-05, | |
| "loss": 0.7429, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 2.2916666666666667e-05, | |
| "loss": 0.7032, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "learning_rate": 2.2685185185185187e-05, | |
| "loss": 0.749, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 2.2453703703703703e-05, | |
| "loss": 0.7122, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.6559, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "learning_rate": 2.1990740740740743e-05, | |
| "loss": 0.5757, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 2.175925925925926e-05, | |
| "loss": 0.5756, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 6.14, | |
| "learning_rate": 2.152777777777778e-05, | |
| "loss": 0.5777, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "learning_rate": 2.1296296296296296e-05, | |
| "loss": 0.5904, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 2.1064814814814816e-05, | |
| "loss": 0.5798, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.5939, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "learning_rate": 2.0601851851851853e-05, | |
| "loss": 0.6095, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.6163, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "learning_rate": 2.013888888888889e-05, | |
| "loss": 0.5634, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 1.990740740740741e-05, | |
| "loss": 0.5909, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "learning_rate": 1.967592592592593e-05, | |
| "loss": 0.5408, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 6.51, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.5435, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 1.9212962962962962e-05, | |
| "loss": 0.5753, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 1.8981481481481482e-05, | |
| "loss": 0.5491, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.5868, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "eval_bleu": 71.6124, | |
| "eval_em": 0.02, | |
| "eval_gen_len": 47.52, | |
| "eval_loss": 0.9307076930999756, | |
| "eval_rm": 0.9556, | |
| "eval_runtime": 330.299, | |
| "eval_samples_per_second": 0.151, | |
| "eval_steps_per_second": 0.021, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.5088, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "learning_rate": 1.8287037037037038e-05, | |
| "loss": 0.6068, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 6.76, | |
| "learning_rate": 1.8055555555555555e-05, | |
| "loss": 0.5327, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 1.7824074074074075e-05, | |
| "loss": 0.5759, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 1.7592592592592595e-05, | |
| "loss": 0.5849, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "learning_rate": 1.736111111111111e-05, | |
| "loss": 0.5648, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 1.712962962962963e-05, | |
| "loss": 0.5632, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 1.6898148148148148e-05, | |
| "loss": 0.5093, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.5295, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 1.6435185185185187e-05, | |
| "loss": 0.433, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "learning_rate": 1.6203703703703704e-05, | |
| "loss": 0.4289, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 1.597222222222222e-05, | |
| "loss": 0.4619, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "learning_rate": 1.574074074074074e-05, | |
| "loss": 0.458, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "learning_rate": 1.550925925925926e-05, | |
| "loss": 0.4155, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 7.26, | |
| "learning_rate": 1.527777777777778e-05, | |
| "loss": 0.4391, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "learning_rate": 1.5046296296296297e-05, | |
| "loss": 0.4264, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 7.34, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.418, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "learning_rate": 1.4583333333333335e-05, | |
| "loss": 0.4285, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "learning_rate": 1.4351851851851853e-05, | |
| "loss": 0.4466, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "learning_rate": 1.412037037037037e-05, | |
| "loss": 0.4499, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "eval_bleu": 77.237, | |
| "eval_em": 0.06, | |
| "eval_gen_len": 46.0, | |
| "eval_loss": 0.8866045475006104, | |
| "eval_rm": 0.9574, | |
| "eval_runtime": 280.5418, | |
| "eval_samples_per_second": 0.178, | |
| "eval_steps_per_second": 0.025, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.4415, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 1.3657407407407408e-05, | |
| "loss": 0.4209, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "learning_rate": 1.3425925925925928e-05, | |
| "loss": 0.4357, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 7.63, | |
| "learning_rate": 1.3194444444444446e-05, | |
| "loss": 0.437, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.4319, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "learning_rate": 1.2731481481481482e-05, | |
| "loss": 0.4578, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.4244, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 1.2268518518518519e-05, | |
| "loss": 0.4046, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "learning_rate": 1.2037037037037037e-05, | |
| "loss": 0.411, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 7.88, | |
| "learning_rate": 1.1805555555555555e-05, | |
| "loss": 0.4219, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 1.1574074074074075e-05, | |
| "loss": 0.3956, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 7.97, | |
| "learning_rate": 1.1342592592592593e-05, | |
| "loss": 0.4333, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.4141, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 8.05, | |
| "learning_rate": 1.087962962962963e-05, | |
| "loss": 0.3199, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 8.09, | |
| "learning_rate": 1.0648148148148148e-05, | |
| "loss": 0.3236, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.13, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 0.3405, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "learning_rate": 1.0185185185185185e-05, | |
| "loss": 0.3411, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 9.953703703703704e-06, | |
| "loss": 0.3166, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "learning_rate": 9.722222222222223e-06, | |
| "loss": 0.3463, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "learning_rate": 9.490740740740741e-06, | |
| "loss": 0.3515, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "eval_bleu": 77.5798, | |
| "eval_em": 0.08, | |
| "eval_gen_len": 47.5, | |
| "eval_loss": 0.9069581627845764, | |
| "eval_rm": 0.9574, | |
| "eval_runtime": 312.8421, | |
| "eval_samples_per_second": 0.16, | |
| "eval_steps_per_second": 0.022, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.3363, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 8.38, | |
| "learning_rate": 9.027777777777777e-06, | |
| "loss": 0.3477, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "learning_rate": 8.796296296296297e-06, | |
| "loss": 0.3454, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 8.46, | |
| "learning_rate": 8.564814814814816e-06, | |
| "loss": 0.3267, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 8.51, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.3406, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "learning_rate": 8.101851851851852e-06, | |
| "loss": 0.3189, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 8.59, | |
| "learning_rate": 7.87037037037037e-06, | |
| "loss": 0.3355, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "learning_rate": 7.63888888888889e-06, | |
| "loss": 0.3315, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.345, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "learning_rate": 7.1759259259259266e-06, | |
| "loss": 0.3332, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.3491, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 6.712962962962964e-06, | |
| "loss": 0.3268, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "learning_rate": 6.481481481481481e-06, | |
| "loss": 0.3328, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 8.88, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.3151, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "learning_rate": 6.0185185185185185e-06, | |
| "loss": 0.3529, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "learning_rate": 5.787037037037038e-06, | |
| "loss": 0.3087, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.3228, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "learning_rate": 5.324074074074074e-06, | |
| "loss": 0.2613, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 5.092592592592592e-06, | |
| "loss": 0.2721, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 4.861111111111111e-06, | |
| "loss": 0.292, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "eval_bleu": 78.649, | |
| "eval_em": 0.06, | |
| "eval_gen_len": 47.96, | |
| "eval_loss": 0.8904868364334106, | |
| "eval_rm": 0.9574, | |
| "eval_runtime": 309.7604, | |
| "eval_samples_per_second": 0.161, | |
| "eval_steps_per_second": 0.023, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "learning_rate": 4.6296296296296296e-06, | |
| "loss": 0.2614, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 9.21, | |
| "learning_rate": 4.398148148148149e-06, | |
| "loss": 0.2522, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 9.25, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.289, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "learning_rate": 3.935185185185185e-06, | |
| "loss": 0.2875, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 9.34, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.2451, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.2855, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 9.42, | |
| "learning_rate": 3.2407407407407406e-06, | |
| "loss": 0.2632, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "learning_rate": 3.0092592592592593e-06, | |
| "loss": 0.2726, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 2.777777777777778e-06, | |
| "loss": 0.2768, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 9.54, | |
| "learning_rate": 2.546296296296296e-06, | |
| "loss": 0.2729, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.59, | |
| "learning_rate": 2.3148148148148148e-06, | |
| "loss": 0.2809, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "loss": 0.2714, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 9.67, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.2693, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 9.71, | |
| "learning_rate": 1.6203703703703703e-06, | |
| "loss": 0.2639, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "learning_rate": 1.388888888888889e-06, | |
| "loss": 0.2667, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "learning_rate": 1.1574074074074074e-06, | |
| "loss": 0.2565, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 9.83, | |
| "learning_rate": 9.259259259259259e-07, | |
| "loss": 0.2855, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "learning_rate": 6.944444444444445e-07, | |
| "loss": 0.2615, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 9.92, | |
| "learning_rate": 4.6296296296296297e-07, | |
| "loss": 0.2709, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 9.96, | |
| "learning_rate": 2.3148148148148148e-07, | |
| "loss": 0.2658, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 9.96, | |
| "eval_bleu": 79.5906, | |
| "eval_em": 0.08, | |
| "eval_gen_len": 48.48, | |
| "eval_loss": 0.9002240896224976, | |
| "eval_rm": 0.9565, | |
| "eval_runtime": 351.0056, | |
| "eval_samples_per_second": 0.142, | |
| "eval_steps_per_second": 0.02, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.2602, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 24100, | |
| "total_flos": 7011428612298048.0, | |
| "train_loss": 0.1476078914309933, | |
| "train_runtime": 5095.148, | |
| "train_samples_per_second": 37.832, | |
| "train_steps_per_second": 4.73 | |
| } | |
| ], | |
| "max_steps": 24100, | |
| "num_train_epochs": 10, | |
| "total_flos": 7011428612298048.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |