| { | |
| "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": 20.4357, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.425e-06, | |
| "loss": 11.928, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.1174999999999999e-05, | |
| "loss": 8.0523, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 1.4925e-05, | |
| "loss": 6.5593, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 1.8675e-05, | |
| "loss": 5.3155, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "eval_loss": 4.558210849761963, | |
| "eval_runtime": 61.1578, | |
| "eval_samples_per_second": 34.256, | |
| "eval_steps_per_second": 4.284, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 2.2424999999999996e-05, | |
| "loss": 4.3699, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 2.6174999999999996e-05, | |
| "loss": 3.7571, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 2.9925e-05, | |
| "loss": 3.5182, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 3.3675e-05, | |
| "loss": 3.4537, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3.7424999999999995e-05, | |
| "loss": 3.3369, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "eval_loss": 3.426917791366577, | |
| "eval_runtime": 62.6973, | |
| "eval_samples_per_second": 33.415, | |
| "eval_steps_per_second": 4.179, | |
| "eval_wer": 1.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "learning_rate": 4.1175e-05, | |
| "loss": 3.2812, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "learning_rate": 4.4924999999999994e-05, | |
| "loss": 3.2389, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "learning_rate": 4.8675e-05, | |
| "loss": 3.1102, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "learning_rate": 5.2424999999999994e-05, | |
| "loss": 2.7161, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 5.6175e-05, | |
| "loss": 2.1785, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "eval_loss": 1.7190676927566528, | |
| "eval_runtime": 62.8007, | |
| "eval_samples_per_second": 33.36, | |
| "eval_steps_per_second": 4.172, | |
| "eval_wer": 0.8831024930747923, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 5.9925e-05, | |
| "loss": 1.9322, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 11.56, | |
| "learning_rate": 6.367499999999999e-05, | |
| "loss": 1.7738, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "learning_rate": 6.7425e-05, | |
| "loss": 1.6853, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.92, | |
| "learning_rate": 7.1175e-05, | |
| "loss": 1.6283, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 7.492499999999999e-05, | |
| "loss": 1.579, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "eval_loss": 1.3603801727294922, | |
| "eval_runtime": 61.857, | |
| "eval_samples_per_second": 33.868, | |
| "eval_steps_per_second": 4.236, | |
| "eval_wer": 0.7647276084949215, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 7.362616822429906e-05, | |
| "loss": 1.5102, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.96, | |
| "learning_rate": 7.222429906542056e-05, | |
| "loss": 1.478, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "learning_rate": 7.082242990654205e-05, | |
| "loss": 1.472, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "learning_rate": 6.942056074766355e-05, | |
| "loss": 1.4333, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "learning_rate": 6.801869158878504e-05, | |
| "loss": 1.3773, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "eval_loss": 1.2737088203430176, | |
| "eval_runtime": 62.7361, | |
| "eval_samples_per_second": 33.394, | |
| "eval_steps_per_second": 4.176, | |
| "eval_wer": 0.7519236688211758, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "learning_rate": 6.661682242990654e-05, | |
| "loss": 1.3823, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 18.37, | |
| "learning_rate": 6.52429906542056e-05, | |
| "loss": 1.3638, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "learning_rate": 6.384112149532709e-05, | |
| "loss": 1.3431, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.73, | |
| "learning_rate": 6.243925233644859e-05, | |
| "loss": 1.3223, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "learning_rate": 6.103738317757008e-05, | |
| "loss": 1.3165, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "eval_loss": 1.245711326599121, | |
| "eval_runtime": 62.7019, | |
| "eval_samples_per_second": 33.412, | |
| "eval_steps_per_second": 4.179, | |
| "eval_wer": 0.740104647583872, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "learning_rate": 5.9635514018691585e-05, | |
| "loss": 1.292, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 21.77, | |
| "learning_rate": 5.823364485981308e-05, | |
| "loss": 1.2607, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 22.45, | |
| "learning_rate": 5.6831775700934575e-05, | |
| "loss": 1.2767, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 23.13, | |
| "learning_rate": 5.542990654205607e-05, | |
| "loss": 1.2518, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 23.81, | |
| "learning_rate": 5.4028037383177566e-05, | |
| "loss": 1.2274, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 23.81, | |
| "eval_loss": 1.3616658449172974, | |
| "eval_runtime": 63.5478, | |
| "eval_samples_per_second": 32.967, | |
| "eval_steps_per_second": 4.123, | |
| "eval_wer": 0.7300707910126193, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 24.49, | |
| "learning_rate": 5.262616822429906e-05, | |
| "loss": 1.1919, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 25.17, | |
| "learning_rate": 5.1224299065420557e-05, | |
| "loss": 1.2453, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 25.85, | |
| "learning_rate": 4.982242990654205e-05, | |
| "loss": 1.2109, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 26.53, | |
| "learning_rate": 4.842056074766355e-05, | |
| "loss": 1.1759, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 27.21, | |
| "learning_rate": 4.701869158878504e-05, | |
| "loss": 1.1787, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.21, | |
| "eval_loss": 1.2067540884017944, | |
| "eval_runtime": 62.6504, | |
| "eval_samples_per_second": 33.44, | |
| "eval_steps_per_second": 4.182, | |
| "eval_wer": 0.7009541397353032, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 27.89, | |
| "learning_rate": 4.561682242990654e-05, | |
| "loss": 1.1783, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 4.421495327102803e-05, | |
| "loss": 1.1709, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 29.25, | |
| "learning_rate": 4.281308411214953e-05, | |
| "loss": 1.1346, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 29.93, | |
| "learning_rate": 4.1411214953271024e-05, | |
| "loss": 1.1589, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 30.61, | |
| "learning_rate": 4.000934579439252e-05, | |
| "loss": 1.1467, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 30.61, | |
| "eval_loss": 1.241578459739685, | |
| "eval_runtime": 62.8423, | |
| "eval_samples_per_second": 33.337, | |
| "eval_steps_per_second": 4.169, | |
| "eval_wer": 0.6946137273007079, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 31.29, | |
| "learning_rate": 3.8621495327102795e-05, | |
| "loss": 1.127, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 31.97, | |
| "learning_rate": 3.72196261682243e-05, | |
| "loss": 1.1027, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 32.65, | |
| "learning_rate": 3.581775700934579e-05, | |
| "loss": 1.125, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 3.4429906542056075e-05, | |
| "loss": 1.1193, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "learning_rate": 3.302803738317757e-05, | |
| "loss": 1.0801, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.01, | |
| "eval_loss": 1.2311749458312988, | |
| "eval_runtime": 61.8112, | |
| "eval_samples_per_second": 33.894, | |
| "eval_steps_per_second": 4.239, | |
| "eval_wer": 0.6990458602646968, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 34.69, | |
| "learning_rate": 3.1626168224299066e-05, | |
| "loss": 1.0891, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 35.37, | |
| "learning_rate": 3.0224299065420558e-05, | |
| "loss": 1.0841, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 36.05, | |
| "learning_rate": 2.8822429906542053e-05, | |
| "loss": 1.0679, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 36.73, | |
| "learning_rate": 2.742056074766355e-05, | |
| "loss": 1.0555, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 37.41, | |
| "learning_rate": 2.6018691588785044e-05, | |
| "loss": 1.0709, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 37.41, | |
| "eval_loss": 1.2984226942062378, | |
| "eval_runtime": 62.056, | |
| "eval_samples_per_second": 33.76, | |
| "eval_steps_per_second": 4.222, | |
| "eval_wer": 0.7137580794090489, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 38.1, | |
| "learning_rate": 2.461682242990654e-05, | |
| "loss": 1.051, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 38.77, | |
| "learning_rate": 2.3214953271028034e-05, | |
| "loss": 1.0309, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 39.46, | |
| "learning_rate": 2.181308411214953e-05, | |
| "loss": 1.0343, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 40.14, | |
| "learning_rate": 2.0411214953271025e-05, | |
| "loss": 1.0456, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 40.81, | |
| "learning_rate": 1.900934579439252e-05, | |
| "loss": 1.0307, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 40.81, | |
| "eval_loss": 1.2048739194869995, | |
| "eval_runtime": 63.482, | |
| "eval_samples_per_second": 33.001, | |
| "eval_steps_per_second": 4.127, | |
| "eval_wer": 0.6871037242228378, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 41.5, | |
| "learning_rate": 1.760747663551402e-05, | |
| "loss": 1.0092, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 42.18, | |
| "learning_rate": 1.6205607476635514e-05, | |
| "loss": 1.0258, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "learning_rate": 1.4803738317757008e-05, | |
| "loss": 1.0129, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 43.54, | |
| "learning_rate": 1.3401869158878503e-05, | |
| "loss": 1.0044, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 44.22, | |
| "learning_rate": 1.1999999999999999e-05, | |
| "loss": 1.0003, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 44.22, | |
| "eval_loss": 1.1955562829971313, | |
| "eval_runtime": 63.8, | |
| "eval_samples_per_second": 32.837, | |
| "eval_steps_per_second": 4.107, | |
| "eval_wer": 0.6841489689135118, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 44.9, | |
| "learning_rate": 1.0598130841121494e-05, | |
| "loss": 1.0068, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 45.58, | |
| "learning_rate": 9.19626168224299e-06, | |
| "loss": 1.0071, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 46.26, | |
| "learning_rate": 7.794392523364485e-06, | |
| "loss": 0.9813, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 46.94, | |
| "learning_rate": 6.392523364485981e-06, | |
| "loss": 0.9888, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 47.62, | |
| "learning_rate": 4.990654205607477e-06, | |
| "loss": 1.004, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 47.62, | |
| "eval_loss": 1.2101075649261475, | |
| "eval_runtime": 62.342, | |
| "eval_samples_per_second": 33.605, | |
| "eval_steps_per_second": 4.203, | |
| "eval_wer": 0.6792859341335795, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 48.3, | |
| "learning_rate": 3.5887850467289714e-06, | |
| "loss": 0.9865, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 48.98, | |
| "learning_rate": 2.186915887850467e-06, | |
| "loss": 0.976, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 49.66, | |
| "learning_rate": 7.850467289719626e-07, | |
| "loss": 1.0041, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 7350, | |
| "total_flos": 2.9231564377949348e+19, | |
| "train_loss": 2.1161629922049388, | |
| "train_runtime": 13636.3379, | |
| "train_samples_per_second": 17.274, | |
| "train_steps_per_second": 0.539 | |
| } | |
| ], | |
| "max_steps": 7350, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.9231564377949348e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |