| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 49.99830220713073, | |
| "global_step": 7350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 3.7125e-06, | |
| "loss": 20.4669, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.425e-06, | |
| "loss": 11.2941, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.1174999999999999e-05, | |
| "loss": 7.683, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 1.4925e-05, | |
| "loss": 6.3219, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 1.8675e-05, | |
| "loss": 5.1844, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "eval_loss": 5.201454162597656, | |
| "eval_runtime": 92.3931, | |
| "eval_samples_per_second": 22.675, | |
| "eval_steps_per_second": 2.836, | |
| "eval_wer": 0.9999454713997492, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 2.2424999999999996e-05, | |
| "loss": 4.3122, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 2.6174999999999996e-05, | |
| "loss": 3.7406, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 2.9925e-05, | |
| "loss": 3.5165, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 3.3675e-05, | |
| "loss": 3.4621, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3.7424999999999995e-05, | |
| "loss": 3.3962, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "eval_loss": 3.401675224304199, | |
| "eval_runtime": 90.8297, | |
| "eval_samples_per_second": 23.065, | |
| "eval_steps_per_second": 2.885, | |
| "eval_wer": 1.0001635858007525, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "learning_rate": 4.1175e-05, | |
| "loss": 3.3137, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "learning_rate": 4.4924999999999994e-05, | |
| "loss": 3.2773, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "learning_rate": 4.8675e-05, | |
| "loss": 3.2224, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "learning_rate": 5.2424999999999994e-05, | |
| "loss": 3.0794, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 5.6175e-05, | |
| "loss": 2.5433, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "eval_loss": 1.688385009765625, | |
| "eval_runtime": 91.7369, | |
| "eval_samples_per_second": 22.837, | |
| "eval_steps_per_second": 2.856, | |
| "eval_wer": 1.0221931403020885, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 5.9925e-05, | |
| "loss": 2.049, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 11.56, | |
| "learning_rate": 6.367499999999999e-05, | |
| "loss": 1.7818, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "learning_rate": 6.7425e-05, | |
| "loss": 1.6648, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "learning_rate": 7.1175e-05, | |
| "loss": 1.5602, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 7.492499999999999e-05, | |
| "loss": 1.5099, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "eval_loss": 0.792931079864502, | |
| "eval_runtime": 92.6306, | |
| "eval_samples_per_second": 22.617, | |
| "eval_steps_per_second": 2.828, | |
| "eval_wer": 1.018757838486286, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 7.362616822429906e-05, | |
| "loss": 1.4211, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.96, | |
| "learning_rate": 7.222429906542056e-05, | |
| "loss": 1.3751, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "learning_rate": 7.083644859813083e-05, | |
| "loss": 1.3544, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "learning_rate": 6.943457943925233e-05, | |
| "loss": 1.3145, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "learning_rate": 6.803271028037383e-05, | |
| "loss": 1.2685, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "eval_loss": 0.6122475862503052, | |
| "eval_runtime": 91.1286, | |
| "eval_samples_per_second": 22.989, | |
| "eval_steps_per_second": 2.875, | |
| "eval_wer": 1.019085010087791, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "learning_rate": 6.663084112149532e-05, | |
| "loss": 1.2627, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 18.37, | |
| "learning_rate": 6.522897196261682e-05, | |
| "loss": 1.2455, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "learning_rate": 6.382710280373831e-05, | |
| "loss": 1.213, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.73, | |
| "learning_rate": 6.243925233644859e-05, | |
| "loss": 1.1857, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "learning_rate": 6.103738317757008e-05, | |
| "loss": 1.1844, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "eval_loss": 0.5433590412139893, | |
| "eval_runtime": 91.9145, | |
| "eval_samples_per_second": 22.793, | |
| "eval_steps_per_second": 2.85, | |
| "eval_wer": 1.019739353290801, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "learning_rate": 5.9635514018691585e-05, | |
| "loss": 1.1482, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 21.77, | |
| "learning_rate": 5.823364485981308e-05, | |
| "loss": 1.1288, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 22.45, | |
| "learning_rate": 5.6831775700934575e-05, | |
| "loss": 1.1429, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 23.13, | |
| "learning_rate": 5.542990654205607e-05, | |
| "loss": 1.1242, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 23.81, | |
| "learning_rate": 5.4028037383177566e-05, | |
| "loss": 1.0945, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 23.81, | |
| "eval_loss": 0.5208417773246765, | |
| "eval_runtime": 89.3768, | |
| "eval_samples_per_second": 23.44, | |
| "eval_steps_per_second": 2.931, | |
| "eval_wer": 1.0315720595452316, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 24.49, | |
| "learning_rate": 5.262616822429906e-05, | |
| "loss": 1.0592, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 25.17, | |
| "learning_rate": 5.1224299065420557e-05, | |
| "loss": 1.1044, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 25.85, | |
| "learning_rate": 4.982242990654205e-05, | |
| "loss": 1.0791, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 26.53, | |
| "learning_rate": 4.842056074766355e-05, | |
| "loss": 1.0414, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 27.21, | |
| "learning_rate": 4.701869158878504e-05, | |
| "loss": 1.0506, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.21, | |
| "eval_loss": 0.49413421750068665, | |
| "eval_runtime": 91.0309, | |
| "eval_samples_per_second": 23.014, | |
| "eval_steps_per_second": 2.878, | |
| "eval_wer": 1.013904793063962, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.89, | |
| "learning_rate": 4.561682242990654e-05, | |
| "loss": 1.051, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 4.421495327102803e-05, | |
| "loss": 1.0389, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 29.25, | |
| "learning_rate": 4.281308411214953e-05, | |
| "loss": 1.0046, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 29.93, | |
| "learning_rate": 4.1411214953271024e-05, | |
| "loss": 1.0228, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 30.61, | |
| "learning_rate": 4.000934579439252e-05, | |
| "loss": 1.0199, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 30.61, | |
| "eval_loss": 0.47357675433158875, | |
| "eval_runtime": 92.1533, | |
| "eval_samples_per_second": 22.734, | |
| "eval_steps_per_second": 2.843, | |
| "eval_wer": 1.0105785484486614, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 31.29, | |
| "learning_rate": 3.8607476635514014e-05, | |
| "loss": 0.9992, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 31.97, | |
| "learning_rate": 3.720560747663551e-05, | |
| "loss": 0.9822, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 32.65, | |
| "learning_rate": 3.5803738317757005e-05, | |
| "loss": 0.9855, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 3.44018691588785e-05, | |
| "loss": 0.9869, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "learning_rate": 3.2999999999999996e-05, | |
| "loss": 0.9546, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "eval_loss": 0.46637728810310364, | |
| "eval_runtime": 92.1156, | |
| "eval_samples_per_second": 22.743, | |
| "eval_steps_per_second": 2.844, | |
| "eval_wer": 1.0163585800752495, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.69, | |
| "learning_rate": 3.159813084112149e-05, | |
| "loss": 0.9592, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 35.37, | |
| "learning_rate": 3.0210280373831774e-05, | |
| "loss": 0.9509, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 36.05, | |
| "learning_rate": 2.880841121495327e-05, | |
| "loss": 0.948, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 36.73, | |
| "learning_rate": 2.7406542056074764e-05, | |
| "loss": 0.9297, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 37.41, | |
| "learning_rate": 2.600467289719626e-05, | |
| "loss": 0.9388, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 37.41, | |
| "eval_loss": 0.4565039277076721, | |
| "eval_runtime": 90.9615, | |
| "eval_samples_per_second": 23.032, | |
| "eval_steps_per_second": 2.88, | |
| "eval_wer": 1.0084519330388788, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 38.1, | |
| "learning_rate": 2.4602803738317755e-05, | |
| "loss": 0.9278, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 38.77, | |
| "learning_rate": 2.3214953271028034e-05, | |
| "loss": 0.9034, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 39.46, | |
| "learning_rate": 2.181308411214953e-05, | |
| "loss": 0.9094, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 40.14, | |
| "learning_rate": 2.0411214953271025e-05, | |
| "loss": 0.9248, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 40.81, | |
| "learning_rate": 1.900934579439252e-05, | |
| "loss": 0.9125, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 40.81, | |
| "eval_loss": 0.4635978043079376, | |
| "eval_runtime": 92.4501, | |
| "eval_samples_per_second": 22.661, | |
| "eval_steps_per_second": 2.834, | |
| "eval_wer": 1.0148317792682262, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 41.5, | |
| "learning_rate": 1.760747663551402e-05, | |
| "loss": 0.8852, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 42.18, | |
| "learning_rate": 1.6205607476635514e-05, | |
| "loss": 0.9073, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "learning_rate": 1.4803738317757008e-05, | |
| "loss": 0.8932, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 43.54, | |
| "learning_rate": 1.3401869158878503e-05, | |
| "loss": 0.8787, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 44.22, | |
| "learning_rate": 1.1999999999999999e-05, | |
| "loss": 0.8733, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 44.22, | |
| "eval_loss": 0.4529880881309509, | |
| "eval_runtime": 92.0867, | |
| "eval_samples_per_second": 22.75, | |
| "eval_steps_per_second": 2.845, | |
| "eval_wer": 1.0153770652707346, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 44.9, | |
| "learning_rate": 1.0598130841121494e-05, | |
| "loss": 0.8866, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 45.58, | |
| "learning_rate": 9.19626168224299e-06, | |
| "loss": 0.8848, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 46.26, | |
| "learning_rate": 7.794392523364485e-06, | |
| "loss": 0.8554, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 46.94, | |
| "learning_rate": 6.392523364485981e-06, | |
| "loss": 0.8712, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 47.62, | |
| "learning_rate": 4.990654205607477e-06, | |
| "loss": 0.8829, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 47.62, | |
| "eval_loss": 0.44942525029182434, | |
| "eval_runtime": 90.8401, | |
| "eval_samples_per_second": 23.062, | |
| "eval_steps_per_second": 2.884, | |
| "eval_wer": 1.015213479469982, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 48.3, | |
| "learning_rate": 3.5887850467289714e-06, | |
| "loss": 0.8635, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 48.98, | |
| "learning_rate": 2.186915887850467e-06, | |
| "loss": 0.8546, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 49.66, | |
| "learning_rate": 7.850467289719626e-07, | |
| "loss": 0.8804, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 7350, | |
| "total_flos": 2.9231469421589975e+19, | |
| "train_loss": 2.0184907983273876, | |
| "train_runtime": 14208.7555, | |
| "train_samples_per_second": 16.578, | |
| "train_steps_per_second": 0.517 | |
| } | |
| ], | |
| "max_steps": 7350, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.9231469421589975e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |