| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 21.08843537414966, | |
| "global_step": 3100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 2.8199999999999998e-05, | |
| "loss": 14.6495, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5.82e-05, | |
| "loss": 8.9314, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_cer": 0.9940322843632806, | |
| "eval_loss": 4.78579044342041, | |
| "eval_runtime": 24.1442, | |
| "eval_samples_per_second": 24.354, | |
| "eval_steps_per_second": 3.065, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.819999999999999e-05, | |
| "loss": 4.2225, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 0.0001182, | |
| "loss": 3.2937, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "eval_cer": 0.9940322843632806, | |
| "eval_loss": 3.2489850521087646, | |
| "eval_runtime": 24.3318, | |
| "eval_samples_per_second": 24.166, | |
| "eval_steps_per_second": 3.041, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.0001482, | |
| "loss": 3.2508, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 0.00017819999999999997, | |
| "loss": 3.2412, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_cer": 0.9940322843632806, | |
| "eval_loss": 3.222844123840332, | |
| "eval_runtime": 24.3579, | |
| "eval_samples_per_second": 24.14, | |
| "eval_steps_per_second": 3.038, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 0.00020819999999999996, | |
| "loss": 3.1855, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 0.0002382, | |
| "loss": 3.0985, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "eval_cer": 0.9940322843632806, | |
| "eval_loss": 3.1458029747009277, | |
| "eval_runtime": 24.7673, | |
| "eval_samples_per_second": 23.741, | |
| "eval_steps_per_second": 2.988, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 0.00026819999999999996, | |
| "loss": 3.0739, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 0.0002982, | |
| "loss": 3.0209, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "eval_cer": 0.9938692320234795, | |
| "eval_loss": 3.004169225692749, | |
| "eval_runtime": 24.3542, | |
| "eval_samples_per_second": 24.144, | |
| "eval_steps_per_second": 3.038, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 0.0002955590551181102, | |
| "loss": 2.9536, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 0.0002908346456692913, | |
| "loss": 2.518, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "eval_cer": 0.5400945703570846, | |
| "eval_loss": 1.770609736442566, | |
| "eval_runtime": 24.8378, | |
| "eval_samples_per_second": 23.674, | |
| "eval_steps_per_second": 2.979, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "learning_rate": 0.00028611023622047245, | |
| "loss": 1.6291, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 0.0002813858267716535, | |
| "loss": 1.2291, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "eval_cer": 0.30572313712701776, | |
| "eval_loss": 0.8816090226173401, | |
| "eval_runtime": 24.4305, | |
| "eval_samples_per_second": 24.068, | |
| "eval_steps_per_second": 3.029, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 0.0002766614173228346, | |
| "loss": 1.0153, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 0.0002719370078740157, | |
| "loss": 0.8879, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "eval_cer": 0.23668677645524214, | |
| "eval_loss": 0.6693652868270874, | |
| "eval_runtime": 24.5154, | |
| "eval_samples_per_second": 23.985, | |
| "eval_steps_per_second": 3.019, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "learning_rate": 0.00026721259842519684, | |
| "loss": 0.8337, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 0.00026248818897637796, | |
| "loss": 0.7717, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "eval_cer": 0.20955486711234306, | |
| "eval_loss": 0.5637524127960205, | |
| "eval_runtime": 24.5794, | |
| "eval_samples_per_second": 23.922, | |
| "eval_steps_per_second": 3.011, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "learning_rate": 0.00025776377952755903, | |
| "loss": 0.6711, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 0.00025303937007874016, | |
| "loss": 0.7003, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "eval_cer": 0.19279308658079244, | |
| "eval_loss": 0.5100580453872681, | |
| "eval_runtime": 24.6049, | |
| "eval_samples_per_second": 23.898, | |
| "eval_steps_per_second": 3.008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 0.0002483149606299212, | |
| "loss": 0.6349, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "learning_rate": 0.00024359055118110235, | |
| "loss": 0.5935, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "eval_cer": 0.17746616663949127, | |
| "eval_loss": 0.46862325072288513, | |
| "eval_runtime": 24.2024, | |
| "eval_samples_per_second": 24.295, | |
| "eval_steps_per_second": 3.058, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "learning_rate": 0.00023886614173228342, | |
| "loss": 0.576, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "learning_rate": 0.00023414173228346455, | |
| "loss": 0.5239, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "eval_cer": 0.1918799934779064, | |
| "eval_loss": 0.4271094799041748, | |
| "eval_runtime": 24.6968, | |
| "eval_samples_per_second": 23.809, | |
| "eval_steps_per_second": 2.996, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "learning_rate": 0.00022941732283464564, | |
| "loss": 0.5207, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "learning_rate": 0.00022469291338582677, | |
| "loss": 0.4855, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.84, | |
| "eval_cer": 0.17550953856187837, | |
| "eval_loss": 0.40729257464408875, | |
| "eval_runtime": 24.885, | |
| "eval_samples_per_second": 23.629, | |
| "eval_steps_per_second": 2.974, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "learning_rate": 0.00021996850393700784, | |
| "loss": 0.4621, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "learning_rate": 0.00021524409448818896, | |
| "loss": 0.4583, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "eval_cer": 0.1760313060492418, | |
| "eval_loss": 0.41417357325553894, | |
| "eval_runtime": 24.8425, | |
| "eval_samples_per_second": 23.669, | |
| "eval_steps_per_second": 2.979, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.86, | |
| "learning_rate": 0.00021051968503937006, | |
| "loss": 0.4415, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 0.00020579527559055118, | |
| "loss": 0.417, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "eval_cer": 0.17156367193869232, | |
| "eval_loss": 0.3833578824996948, | |
| "eval_runtime": 24.455, | |
| "eval_samples_per_second": 24.044, | |
| "eval_steps_per_second": 3.026, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "learning_rate": 0.00020107086614173225, | |
| "loss": 0.3959, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 0.00019634645669291338, | |
| "loss": 0.4074, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "eval_cer": 0.16301972933311593, | |
| "eval_loss": 0.3626195192337036, | |
| "eval_runtime": 24.5184, | |
| "eval_samples_per_second": 23.982, | |
| "eval_steps_per_second": 3.018, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 11.22, | |
| "learning_rate": 0.00019162204724409448, | |
| "loss": 0.3584, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 11.56, | |
| "learning_rate": 0.0001868976377952756, | |
| "loss": 0.3682, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 11.56, | |
| "eval_cer": 0.15685635088863525, | |
| "eval_loss": 0.36927542090415955, | |
| "eval_runtime": 24.3751, | |
| "eval_samples_per_second": 24.123, | |
| "eval_steps_per_second": 3.036, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "learning_rate": 0.00018217322834645667, | |
| "loss": 0.3455, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "learning_rate": 0.00017744881889763777, | |
| "loss": 0.3245, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "eval_cer": 0.15816076960704387, | |
| "eval_loss": 0.3740461468696594, | |
| "eval_runtime": 24.7162, | |
| "eval_samples_per_second": 23.79, | |
| "eval_steps_per_second": 2.994, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.59, | |
| "learning_rate": 0.0001727244094488189, | |
| "loss": 0.3208, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "learning_rate": 0.000168, | |
| "loss": 0.3063, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "eval_cer": 0.15904125224196966, | |
| "eval_loss": 0.3622555434703827, | |
| "eval_runtime": 24.4729, | |
| "eval_samples_per_second": 24.027, | |
| "eval_steps_per_second": 3.024, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 13.27, | |
| "learning_rate": 0.0001632755905511811, | |
| "loss": 0.3019, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 13.61, | |
| "learning_rate": 0.00015855118110236219, | |
| "loss": 0.2945, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.61, | |
| "eval_cer": 0.16634599706505787, | |
| "eval_loss": 0.3725011348724365, | |
| "eval_runtime": 25.0023, | |
| "eval_samples_per_second": 23.518, | |
| "eval_steps_per_second": 2.96, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.95, | |
| "learning_rate": 0.0001538267716535433, | |
| "loss": 0.279, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 0.0001491023622047244, | |
| "loss": 0.2674, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "eval_cer": 0.15731289744007826, | |
| "eval_loss": 0.3531067371368408, | |
| "eval_runtime": 24.8381, | |
| "eval_samples_per_second": 23.673, | |
| "eval_steps_per_second": 2.979, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "learning_rate": 0.0001443779527559055, | |
| "loss": 0.2584, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 14.97, | |
| "learning_rate": 0.0001396535433070866, | |
| "loss": 0.2796, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 14.97, | |
| "eval_cer": 0.14808413500733736, | |
| "eval_loss": 0.3606802523136139, | |
| "eval_runtime": 24.8151, | |
| "eval_samples_per_second": 23.695, | |
| "eval_steps_per_second": 2.982, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 15.31, | |
| "learning_rate": 0.0001349291338582677, | |
| "loss": 0.2462, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "learning_rate": 0.0001302047244094488, | |
| "loss": 0.256, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "eval_cer": 0.15819338007500408, | |
| "eval_loss": 0.3580550253391266, | |
| "eval_runtime": 24.5695, | |
| "eval_samples_per_second": 23.932, | |
| "eval_steps_per_second": 3.012, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 15.99, | |
| "learning_rate": 0.00012548031496062992, | |
| "loss": 0.2524, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "learning_rate": 0.00012075590551181102, | |
| "loss": 0.2219, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.33, | |
| "eval_cer": 0.14801891407141693, | |
| "eval_loss": 0.35925593972206116, | |
| "eval_runtime": 24.982, | |
| "eval_samples_per_second": 23.537, | |
| "eval_steps_per_second": 2.962, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 0.0001160314960629921, | |
| "loss": 0.2364, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "learning_rate": 0.00011130708661417321, | |
| "loss": 0.2291, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "eval_cer": 0.1471058209685309, | |
| "eval_loss": 0.35567909479141235, | |
| "eval_runtime": 24.4749, | |
| "eval_samples_per_second": 24.025, | |
| "eval_steps_per_second": 3.024, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 17.35, | |
| "learning_rate": 0.00010658267716535431, | |
| "loss": 0.2045, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "learning_rate": 0.00010185826771653542, | |
| "loss": 0.2172, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "eval_cer": 0.14792108266753629, | |
| "eval_loss": 0.3606509566307068, | |
| "eval_runtime": 25.1105, | |
| "eval_samples_per_second": 23.416, | |
| "eval_steps_per_second": 2.947, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 18.03, | |
| "learning_rate": 9.713385826771652e-05, | |
| "loss": 0.2271, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 18.37, | |
| "learning_rate": 9.240944881889763e-05, | |
| "loss": 0.1858, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 18.37, | |
| "eval_cer": 0.15144301320723952, | |
| "eval_loss": 0.3589307963848114, | |
| "eval_runtime": 24.5005, | |
| "eval_samples_per_second": 24.0, | |
| "eval_steps_per_second": 3.02, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 18.71, | |
| "learning_rate": 8.768503937007873e-05, | |
| "loss": 0.1995, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "learning_rate": 8.296062992125984e-05, | |
| "loss": 0.1872, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "eval_cer": 0.14766019892385457, | |
| "eval_loss": 0.36663514375686646, | |
| "eval_runtime": 24.4383, | |
| "eval_samples_per_second": 24.061, | |
| "eval_steps_per_second": 3.028, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 19.39, | |
| "learning_rate": 7.823622047244094e-05, | |
| "loss": 0.177, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 19.73, | |
| "learning_rate": 7.351181102362205e-05, | |
| "loss": 0.1855, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 19.73, | |
| "eval_cer": 0.1432577857492255, | |
| "eval_loss": 0.3651330769062042, | |
| "eval_runtime": 24.7826, | |
| "eval_samples_per_second": 23.726, | |
| "eval_steps_per_second": 2.986, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 20.07, | |
| "learning_rate": 6.878740157480315e-05, | |
| "loss": 0.1773, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "learning_rate": 6.406299212598424e-05, | |
| "loss": 0.185, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.41, | |
| "eval_cer": 0.1447578672753954, | |
| "eval_loss": 0.36554473638534546, | |
| "eval_runtime": 24.7061, | |
| "eval_samples_per_second": 23.8, | |
| "eval_steps_per_second": 2.995, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "learning_rate": 5.933858267716535e-05, | |
| "loss": 0.1735, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "learning_rate": 5.461417322834645e-05, | |
| "loss": 0.1599, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 21.09, | |
| "eval_cer": 0.14867112343062122, | |
| "eval_loss": 0.37344449758529663, | |
| "eval_runtime": 24.8752, | |
| "eval_samples_per_second": 23.638, | |
| "eval_steps_per_second": 2.975, | |
| "step": 3100 | |
| } | |
| ], | |
| "max_steps": 3675, | |
| "num_train_epochs": 25, | |
| "total_flos": 1.2047450649580579e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |