| { | |
| "best_metric": 0.2257038652896881, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-swagen-combined-30hrs-model/checkpoint-2900", | |
| "epoch": 1.315852442671984, | |
| "eval_steps": 100, | |
| "global_step": 3300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03988035892323031, | |
| "grad_norm": 51.95917510986328, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 16.9208, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03988035892323031, | |
| "eval_loss": 4.268084526062012, | |
| "eval_runtime": 62.3338, | |
| "eval_samples_per_second": 18.16, | |
| "eval_steps_per_second": 4.54, | |
| "eval_wer": 0.9999006754072308, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07976071784646062, | |
| "grad_norm": 27.331029891967773, | |
| "learning_rate": 0.00029962055651710823, | |
| "loss": 7.2021, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07976071784646062, | |
| "eval_loss": 3.4067752361297607, | |
| "eval_runtime": 62.1258, | |
| "eval_samples_per_second": 18.221, | |
| "eval_steps_per_second": 4.555, | |
| "eval_wer": 1.0026817640047676, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11964107676969092, | |
| "grad_norm": 32.389076232910156, | |
| "learning_rate": 0.0002992211423245906, | |
| "loss": 6.725, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.11964107676969092, | |
| "eval_loss": 3.1580252647399902, | |
| "eval_runtime": 62.3765, | |
| "eval_samples_per_second": 18.148, | |
| "eval_steps_per_second": 4.537, | |
| "eval_wer": 1.014600715137068, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15952143569292124, | |
| "grad_norm": 34.745182037353516, | |
| "learning_rate": 0.0002988217281320729, | |
| "loss": 6.2214, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.15952143569292124, | |
| "eval_loss": 3.043919324874878, | |
| "eval_runtime": 62.1466, | |
| "eval_samples_per_second": 18.215, | |
| "eval_steps_per_second": 4.554, | |
| "eval_wer": 1.0026817640047676, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.19940179461615154, | |
| "grad_norm": 25.307340621948242, | |
| "learning_rate": 0.0002984223139395553, | |
| "loss": 5.9996, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.19940179461615154, | |
| "eval_loss": 3.004955768585205, | |
| "eval_runtime": 62.9799, | |
| "eval_samples_per_second": 17.974, | |
| "eval_steps_per_second": 4.493, | |
| "eval_wer": 1.0015891934843066, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23928215353938184, | |
| "grad_norm": 6.2275800704956055, | |
| "learning_rate": 0.00029802289974703767, | |
| "loss": 5.8133, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.23928215353938184, | |
| "eval_loss": 2.9414522647857666, | |
| "eval_runtime": 62.5714, | |
| "eval_samples_per_second": 18.091, | |
| "eval_steps_per_second": 4.523, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.27916251246261214, | |
| "grad_norm": 6.14309549331665, | |
| "learning_rate": 0.00029762348555452, | |
| "loss": 5.805, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.27916251246261214, | |
| "eval_loss": 2.8951313495635986, | |
| "eval_runtime": 62.4303, | |
| "eval_samples_per_second": 18.132, | |
| "eval_steps_per_second": 4.533, | |
| "eval_wer": 0.9997020262216925, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3190428713858425, | |
| "grad_norm": 14.912216186523438, | |
| "learning_rate": 0.00029722407136200237, | |
| "loss": 5.7621, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3190428713858425, | |
| "eval_loss": 2.905362844467163, | |
| "eval_runtime": 62.3651, | |
| "eval_samples_per_second": 18.151, | |
| "eval_steps_per_second": 4.538, | |
| "eval_wer": 0.9996027016289233, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3589232303090728, | |
| "grad_norm": 3.54911470413208, | |
| "learning_rate": 0.00029682465716948474, | |
| "loss": 5.6973, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3589232303090728, | |
| "eval_loss": 2.810933828353882, | |
| "eval_runtime": 63.0608, | |
| "eval_samples_per_second": 17.951, | |
| "eval_steps_per_second": 4.488, | |
| "eval_wer": 0.9908621374652364, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3988035892323031, | |
| "grad_norm": 5.402283668518066, | |
| "learning_rate": 0.00029642524297696706, | |
| "loss": 5.6541, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3988035892323031, | |
| "eval_loss": 2.835520029067993, | |
| "eval_runtime": 62.7046, | |
| "eval_samples_per_second": 18.053, | |
| "eval_steps_per_second": 4.513, | |
| "eval_wer": 0.9910607866507747, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4386839481555334, | |
| "grad_norm": 5.038330554962158, | |
| "learning_rate": 0.00029602582878444943, | |
| "loss": 5.4159, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4386839481555334, | |
| "eval_loss": 2.7631733417510986, | |
| "eval_runtime": 62.482, | |
| "eval_samples_per_second": 18.117, | |
| "eval_steps_per_second": 4.529, | |
| "eval_wer": 0.978347238776321, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4785643070787637, | |
| "grad_norm": 10.235861778259277, | |
| "learning_rate": 0.0002956264145919318, | |
| "loss": 5.4112, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.4785643070787637, | |
| "eval_loss": 2.6498703956604004, | |
| "eval_runtime": 62.6482, | |
| "eval_samples_per_second": 18.069, | |
| "eval_steps_per_second": 4.517, | |
| "eval_wer": 0.9793404847040127, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.518444666001994, | |
| "grad_norm": 15.807747840881348, | |
| "learning_rate": 0.0002952270003994142, | |
| "loss": 4.2059, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.518444666001994, | |
| "eval_loss": 0.36752405762672424, | |
| "eval_runtime": 63.4538, | |
| "eval_samples_per_second": 17.84, | |
| "eval_steps_per_second": 4.46, | |
| "eval_wer": 0.261720301946762, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5583250249252243, | |
| "grad_norm": 3.0918314456939697, | |
| "learning_rate": 0.0002948275862068965, | |
| "loss": 0.6493, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5583250249252243, | |
| "eval_loss": 0.27466678619384766, | |
| "eval_runtime": 62.8469, | |
| "eval_samples_per_second": 18.012, | |
| "eval_steps_per_second": 4.503, | |
| "eval_wer": 0.20768772348033374, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5982053838484547, | |
| "grad_norm": 3.152963638305664, | |
| "learning_rate": 0.0002944281720143789, | |
| "loss": 0.5624, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.5982053838484547, | |
| "eval_loss": 0.2648875117301941, | |
| "eval_runtime": 63.2142, | |
| "eval_samples_per_second": 17.907, | |
| "eval_steps_per_second": 4.477, | |
| "eval_wer": 0.20232419547079858, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.638085742771685, | |
| "grad_norm": 2.553431749343872, | |
| "learning_rate": 0.00029402875782186125, | |
| "loss": 0.5197, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.638085742771685, | |
| "eval_loss": 0.2619471848011017, | |
| "eval_runtime": 63.2314, | |
| "eval_samples_per_second": 17.902, | |
| "eval_steps_per_second": 4.476, | |
| "eval_wer": 0.19884783472387763, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "grad_norm": 11.26433277130127, | |
| "learning_rate": 0.00029363333777126876, | |
| "loss": 0.4715, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "eval_loss": 0.2589423656463623, | |
| "eval_runtime": 63.8104, | |
| "eval_samples_per_second": 17.74, | |
| "eval_steps_per_second": 4.435, | |
| "eval_wer": 0.19815256257449346, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7178464606181456, | |
| "grad_norm": 2.320037603378296, | |
| "learning_rate": 0.00029323392357875113, | |
| "loss": 0.5126, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7178464606181456, | |
| "eval_loss": 0.2518182694911957, | |
| "eval_runtime": 63.1237, | |
| "eval_samples_per_second": 17.933, | |
| "eval_steps_per_second": 4.483, | |
| "eval_wer": 0.19785458879618595, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7577268195413759, | |
| "grad_norm": 2.7758800983428955, | |
| "learning_rate": 0.0002928345093862335, | |
| "loss": 0.4916, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7577268195413759, | |
| "eval_loss": 0.254902720451355, | |
| "eval_runtime": 63.0039, | |
| "eval_samples_per_second": 17.967, | |
| "eval_steps_per_second": 4.492, | |
| "eval_wer": 0.19576877234803336, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7976071784646062, | |
| "grad_norm": 3.670294761657715, | |
| "learning_rate": 0.0002924350951937158, | |
| "loss": 0.4667, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7976071784646062, | |
| "eval_loss": 0.2501005530357361, | |
| "eval_runtime": 63.1402, | |
| "eval_samples_per_second": 17.928, | |
| "eval_steps_per_second": 4.482, | |
| "eval_wer": 0.1946762018275725, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8374875373878365, | |
| "grad_norm": 3.5203053951263428, | |
| "learning_rate": 0.0002920356810011982, | |
| "loss": 0.4713, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8374875373878365, | |
| "eval_loss": 0.24789908528327942, | |
| "eval_runtime": 63.6613, | |
| "eval_samples_per_second": 17.782, | |
| "eval_steps_per_second": 4.445, | |
| "eval_wer": 0.19427890345649582, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8773678963110668, | |
| "grad_norm": 3.342277765274048, | |
| "learning_rate": 0.0002916362668086806, | |
| "loss": 0.4875, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.8773678963110668, | |
| "eval_loss": 0.24493974447250366, | |
| "eval_runtime": 63.0142, | |
| "eval_samples_per_second": 17.964, | |
| "eval_steps_per_second": 4.491, | |
| "eval_wer": 0.19308700834326578, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.9172482552342971, | |
| "grad_norm": 4.0236711502075195, | |
| "learning_rate": 0.0002912368526161629, | |
| "loss": 0.4611, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.9172482552342971, | |
| "eval_loss": 0.24361559748649597, | |
| "eval_runtime": 62.9448, | |
| "eval_samples_per_second": 17.984, | |
| "eval_steps_per_second": 4.496, | |
| "eval_wer": 0.19348430671434247, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.9571286141575274, | |
| "grad_norm": 2.9615795612335205, | |
| "learning_rate": 0.00029083743842364527, | |
| "loss": 0.4587, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9571286141575274, | |
| "eval_loss": 0.24336393177509308, | |
| "eval_runtime": 63.2457, | |
| "eval_samples_per_second": 17.898, | |
| "eval_steps_per_second": 4.475, | |
| "eval_wer": 0.19278903456495827, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9970089730807578, | |
| "grad_norm": 2.3204922676086426, | |
| "learning_rate": 0.00029043802423112764, | |
| "loss": 0.4679, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.9970089730807578, | |
| "eval_loss": 0.24085378646850586, | |
| "eval_runtime": 63.8549, | |
| "eval_samples_per_second": 17.728, | |
| "eval_steps_per_second": 4.432, | |
| "eval_wer": 0.18951132300357568, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.036689930209372, | |
| "grad_norm": 1.1865150928497314, | |
| "learning_rate": 0.00029003861003861, | |
| "loss": 0.4141, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.036689930209372, | |
| "eval_loss": 0.23312124609947205, | |
| "eval_runtime": 63.2647, | |
| "eval_samples_per_second": 17.893, | |
| "eval_steps_per_second": 4.473, | |
| "eval_wer": 0.18961064759634486, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.076570289132602, | |
| "grad_norm": 1.3028316497802734, | |
| "learning_rate": 0.0002896391958460924, | |
| "loss": 0.4263, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.076570289132602, | |
| "eval_loss": 0.23293058574199677, | |
| "eval_runtime": 62.8596, | |
| "eval_samples_per_second": 18.008, | |
| "eval_steps_per_second": 4.502, | |
| "eval_wer": 0.19199443782280493, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.1164506480558325, | |
| "grad_norm": 7.91491174697876, | |
| "learning_rate": 0.0002892397816535747, | |
| "loss": 0.4142, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.1164506480558325, | |
| "eval_loss": 0.23239396512508392, | |
| "eval_runtime": 63.0701, | |
| "eval_samples_per_second": 17.948, | |
| "eval_steps_per_second": 4.487, | |
| "eval_wer": 0.1917957886372666, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.1563310069790629, | |
| "grad_norm": 1.7904026508331299, | |
| "learning_rate": 0.0002888403674610571, | |
| "loss": 0.4606, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.1563310069790629, | |
| "eval_loss": 0.2257038652896881, | |
| "eval_runtime": 63.835, | |
| "eval_samples_per_second": 17.733, | |
| "eval_steps_per_second": 4.433, | |
| "eval_wer": 0.19427890345649582, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.196211365902293, | |
| "grad_norm": 2.778794527053833, | |
| "learning_rate": 0.00028844095326853946, | |
| "loss": 0.4048, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.196211365902293, | |
| "eval_loss": 0.2288675457239151, | |
| "eval_runtime": 63.4733, | |
| "eval_samples_per_second": 17.834, | |
| "eval_steps_per_second": 4.459, | |
| "eval_wer": 0.19278903456495827, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2360917248255234, | |
| "grad_norm": 1.2311575412750244, | |
| "learning_rate": 0.00028804153907602183, | |
| "loss": 0.4172, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.2360917248255234, | |
| "eval_loss": 0.2326343059539795, | |
| "eval_runtime": 63.2571, | |
| "eval_samples_per_second": 17.895, | |
| "eval_steps_per_second": 4.474, | |
| "eval_wer": 0.19378228049264998, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.2759720837487538, | |
| "grad_norm": 1.667809247970581, | |
| "learning_rate": 0.00028764212488350415, | |
| "loss": 0.4294, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.2759720837487538, | |
| "eval_loss": 0.232680082321167, | |
| "eval_runtime": 63.7285, | |
| "eval_samples_per_second": 17.763, | |
| "eval_steps_per_second": 4.441, | |
| "eval_wer": 0.1940802542709575, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.315852442671984, | |
| "grad_norm": 2.067225456237793, | |
| "learning_rate": 0.00028724271069098653, | |
| "loss": 0.468, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.315852442671984, | |
| "eval_loss": 0.22773675620555878, | |
| "eval_runtime": 63.9749, | |
| "eval_samples_per_second": 17.694, | |
| "eval_steps_per_second": 4.424, | |
| "eval_wer": 0.19219308700834326, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.315852442671984, | |
| "step": 3300, | |
| "total_flos": 1.351677187667784e+19, | |
| "train_loss": 2.917839215596517, | |
| "train_runtime": 6068.548, | |
| "train_samples_per_second": 99.162, | |
| "train_steps_per_second": 12.393 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 75210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.351677187667784e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |