| { | |
| "best_metric": 3.2635854592980795, | |
| "best_model_checkpoint": "OUTCOMESAI/whisper-large-v3-medical/checkpoint-5000", | |
| "epoch": 3.8255547054322876, | |
| "eval_steps": 200, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07651109410864575, | |
| "grad_norm": 4.125481605529785, | |
| "learning_rate": 4.2874883363489934e-07, | |
| "loss": 6.0929, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1530221882172915, | |
| "grad_norm": 4.159045696258545, | |
| "learning_rate": 4.971255772939331e-07, | |
| "loss": 4.2439, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1530221882172915, | |
| "eval_loss": 0.29345703125, | |
| "eval_runtime": 988.7646, | |
| "eval_samples_per_second": 1.131, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.507771356816114, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22953328232593725, | |
| "grad_norm": 2.488393545150757, | |
| "learning_rate": 4.903124999999999e-07, | |
| "loss": 3.5952, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.306044376434583, | |
| "grad_norm": 2.386241912841797, | |
| "learning_rate": 4.798958333333334e-07, | |
| "loss": 3.3374, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.306044376434583, | |
| "eval_loss": 0.2734375, | |
| "eval_runtime": 1214.5433, | |
| "eval_samples_per_second": 0.921, | |
| "eval_steps_per_second": 0.029, | |
| "eval_wer": 4.696129057058288, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.38255547054322875, | |
| "grad_norm": 2.348032236099243, | |
| "learning_rate": 4.6947916666666664e-07, | |
| "loss": 3.1968, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4590665646518745, | |
| "grad_norm": 2.6516611576080322, | |
| "learning_rate": 4.5906249999999995e-07, | |
| "loss": 3.0833, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4590665646518745, | |
| "eval_loss": 0.267333984375, | |
| "eval_runtime": 991.4442, | |
| "eval_samples_per_second": 1.128, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.273285240188102, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5355776587605203, | |
| "grad_norm": 3.4489781856536865, | |
| "learning_rate": 4.486458333333333e-07, | |
| "loss": 2.6306, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.612088752869166, | |
| "grad_norm": 2.9474337100982666, | |
| "learning_rate": 4.3822916666666667e-07, | |
| "loss": 1.8243, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.612088752869166, | |
| "eval_loss": 0.26806640625, | |
| "eval_runtime": 991.8395, | |
| "eval_samples_per_second": 1.127, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.4372973873377495, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6885998469778117, | |
| "grad_norm": 2.7142157554626465, | |
| "learning_rate": 4.278125e-07, | |
| "loss": 1.382, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7651109410864575, | |
| "grad_norm": 2.386582612991333, | |
| "learning_rate": 4.1739583333333334e-07, | |
| "loss": 1.1288, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7651109410864575, | |
| "eval_loss": 0.2548828125, | |
| "eval_runtime": 991.8205, | |
| "eval_samples_per_second": 1.127, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.277129274886922, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8416220351951033, | |
| "grad_norm": 2.2802257537841797, | |
| "learning_rate": 4.0697916666666665e-07, | |
| "loss": 0.9419, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.918133129303749, | |
| "grad_norm": 2.0405514240264893, | |
| "learning_rate": 3.9656249999999996e-07, | |
| "loss": 0.8199, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.918133129303749, | |
| "eval_loss": 0.2412109375, | |
| "eval_runtime": 994.3546, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.204092615609343, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9946442234123948, | |
| "grad_norm": 1.88164222240448, | |
| "learning_rate": 3.861458333333333e-07, | |
| "loss": 0.7493, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.0711553175210407, | |
| "grad_norm": 1.8802990913391113, | |
| "learning_rate": 3.7572916666666663e-07, | |
| "loss": 0.681, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.0711553175210407, | |
| "eval_loss": 0.2310791015625, | |
| "eval_runtime": 990.487, | |
| "eval_samples_per_second": 1.129, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.105429058339634, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1476664116296864, | |
| "grad_norm": 1.6826539039611816, | |
| "learning_rate": 3.6531249999999994e-07, | |
| "loss": 0.6296, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2241775057383322, | |
| "grad_norm": 1.7766120433807373, | |
| "learning_rate": 3.5489583333333336e-07, | |
| "loss": 0.5798, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.2241775057383322, | |
| "eval_loss": 0.21923828125, | |
| "eval_runtime": 989.3653, | |
| "eval_samples_per_second": 1.13, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 4.009328190869136, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.300688599846978, | |
| "grad_norm": 1.597964882850647, | |
| "learning_rate": 3.4447916666666667e-07, | |
| "loss": 0.5603, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3771996939556237, | |
| "grad_norm": 1.5654646158218384, | |
| "learning_rate": 3.340625e-07, | |
| "loss": 0.5233, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.3771996939556237, | |
| "eval_loss": 0.2071533203125, | |
| "eval_runtime": 987.6154, | |
| "eval_samples_per_second": 1.132, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.8927258050049334, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.4537107880642695, | |
| "grad_norm": 1.5115400552749634, | |
| "learning_rate": 3.2364583333333334e-07, | |
| "loss": 0.4919, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.5302218821729152, | |
| "grad_norm": 1.5361956357955933, | |
| "learning_rate": 3.1322916666666665e-07, | |
| "loss": 0.463, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5302218821729152, | |
| "eval_loss": 0.19921875, | |
| "eval_runtime": 984.868, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 3.8196891457273554, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.606732976281561, | |
| "grad_norm": 1.4659453630447388, | |
| "learning_rate": 3.0281249999999996e-07, | |
| "loss": 0.4509, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.6832440703902067, | |
| "grad_norm": 1.384372591972351, | |
| "learning_rate": 2.923958333333333e-07, | |
| "loss": 0.428, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.6832440703902067, | |
| "eval_loss": 0.195068359375, | |
| "eval_runtime": 985.7714, | |
| "eval_samples_per_second": 1.134, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 3.7748420742411235, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.7597551644988525, | |
| "grad_norm": 1.451076865196228, | |
| "learning_rate": 2.8197916666666663e-07, | |
| "loss": 0.4161, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.836266258607498, | |
| "grad_norm": 1.5242278575897217, | |
| "learning_rate": 2.715625e-07, | |
| "loss": 0.3944, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.836266258607498, | |
| "eval_loss": 0.1866455078125, | |
| "eval_runtime": 987.2353, | |
| "eval_samples_per_second": 1.132, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.67745986187102, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.9127773527161438, | |
| "grad_norm": 1.31293785572052, | |
| "learning_rate": 2.6114583333333336e-07, | |
| "loss": 0.38, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9892884468247896, | |
| "grad_norm": 1.3785984516143799, | |
| "learning_rate": 2.5072916666666667e-07, | |
| "loss": 0.3682, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.9892884468247896, | |
| "eval_loss": 0.17919921875, | |
| "eval_runtime": 989.9094, | |
| "eval_samples_per_second": 1.129, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.604423202593442, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.0657995409334355, | |
| "grad_norm": 1.3718817234039307, | |
| "learning_rate": 2.403125e-07, | |
| "loss": 0.36, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.1423106350420813, | |
| "grad_norm": 1.3219904899597168, | |
| "learning_rate": 2.298958333333333e-07, | |
| "loss": 0.3543, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.1423106350420813, | |
| "eval_loss": 0.1724853515625, | |
| "eval_runtime": 995.0367, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.530105198416258, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.218821729150727, | |
| "grad_norm": 1.2633302211761475, | |
| "learning_rate": 2.1947916666666665e-07, | |
| "loss": 0.3438, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.295332823259373, | |
| "grad_norm": 1.2741142511367798, | |
| "learning_rate": 2.090625e-07, | |
| "loss": 0.3368, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.295332823259373, | |
| "eval_loss": 0.17138671875, | |
| "eval_runtime": 991.227, | |
| "eval_samples_per_second": 1.128, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.490383506528452, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.371843917368018, | |
| "grad_norm": 1.2177200317382812, | |
| "learning_rate": 1.9864583333333332e-07, | |
| "loss": 0.3218, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.4483550114766643, | |
| "grad_norm": 1.24095618724823, | |
| "learning_rate": 1.8822916666666666e-07, | |
| "loss": 0.3136, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.4483550114766643, | |
| "eval_loss": 0.164794921875, | |
| "eval_runtime": 995.8265, | |
| "eval_samples_per_second": 1.123, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.45706853913868, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.5248661055853097, | |
| "grad_norm": 1.146142840385437, | |
| "learning_rate": 1.778125e-07, | |
| "loss": 0.3269, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.601377199693956, | |
| "grad_norm": 1.1721028089523315, | |
| "learning_rate": 1.6739583333333333e-07, | |
| "loss": 0.3121, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.601377199693956, | |
| "eval_loss": 0.160400390625, | |
| "eval_runtime": 990.5861, | |
| "eval_samples_per_second": 1.129, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.4237535717489074, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.677888293802601, | |
| "grad_norm": 1.5002176761627197, | |
| "learning_rate": 1.5697916666666666e-07, | |
| "loss": 0.2999, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.7543993879112474, | |
| "grad_norm": 1.250117301940918, | |
| "learning_rate": 1.465625e-07, | |
| "loss": 0.2959, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.7543993879112474, | |
| "eval_loss": 0.1561279296875, | |
| "eval_runtime": 997.6176, | |
| "eval_samples_per_second": 1.121, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.395563983957562, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.8309104820198927, | |
| "grad_norm": 1.2006161212921143, | |
| "learning_rate": 1.361458333333333e-07, | |
| "loss": 0.2942, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.907421576128539, | |
| "grad_norm": 1.2639755010604858, | |
| "learning_rate": 1.2572916666666667e-07, | |
| "loss": 0.2912, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.907421576128539, | |
| "eval_loss": 0.15380859375, | |
| "eval_runtime": 989.6556, | |
| "eval_samples_per_second": 1.13, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.3737811206642494, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.9839326702371842, | |
| "grad_norm": 1.1383545398712158, | |
| "learning_rate": 1.153125e-07, | |
| "loss": 0.2921, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.06044376434583, | |
| "grad_norm": 1.1553541421890259, | |
| "learning_rate": 1.0489583333333332e-07, | |
| "loss": 0.2767, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.06044376434583, | |
| "eval_loss": 0.151123046875, | |
| "eval_runtime": 994.8808, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.3455915328729033, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.1369548584544757, | |
| "grad_norm": 14.852476119995117, | |
| "learning_rate": 9.447916666666667e-08, | |
| "loss": 0.2857, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 3.2134659525631215, | |
| "grad_norm": 3.1388800144195557, | |
| "learning_rate": 8.406249999999999e-08, | |
| "loss": 0.2848, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.2134659525631215, | |
| "eval_loss": 0.148681640625, | |
| "eval_runtime": 1017.2976, | |
| "eval_samples_per_second": 1.099, | |
| "eval_steps_per_second": 0.034, | |
| "eval_wer": 3.319964634880771, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.2899770466717673, | |
| "grad_norm": 1.152143120765686, | |
| "learning_rate": 7.364583333333333e-08, | |
| "loss": 0.2756, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 3.366488140780413, | |
| "grad_norm": 1.1069058179855347, | |
| "learning_rate": 6.322916666666666e-08, | |
| "loss": 0.274, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 3.366488140780413, | |
| "eval_loss": 0.1474609375, | |
| "eval_runtime": 993.2425, | |
| "eval_samples_per_second": 1.126, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.284086977691785, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 3.442999234889059, | |
| "grad_norm": 1.054897427558899, | |
| "learning_rate": 5.291666666666667e-08, | |
| "loss": 0.2729, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.5195103289977046, | |
| "grad_norm": 1.0845283269882202, | |
| "learning_rate": 4.2500000000000003e-08, | |
| "loss": 0.2694, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.5195103289977046, | |
| "eval_loss": 0.1463623046875, | |
| "eval_runtime": 1016.9332, | |
| "eval_samples_per_second": 1.099, | |
| "eval_steps_per_second": 0.034, | |
| "eval_wer": 3.282805632792179, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.5960214231063503, | |
| "grad_norm": 1.0754883289337158, | |
| "learning_rate": 3.208333333333333e-08, | |
| "loss": 0.2635, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 3.672532517214996, | |
| "grad_norm": 1.3197473287582397, | |
| "learning_rate": 2.1770833333333332e-08, | |
| "loss": 0.2731, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.672532517214996, | |
| "eval_loss": 0.1455078125, | |
| "eval_runtime": 989.4621, | |
| "eval_samples_per_second": 1.13, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.268710838896506, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.749043611323642, | |
| "grad_norm": 1.156948208808899, | |
| "learning_rate": 1.1354166666666667e-08, | |
| "loss": 0.2589, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 3.8255547054322876, | |
| "grad_norm": 1.1573657989501953, | |
| "learning_rate": 9.375e-10, | |
| "loss": 0.2677, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.8255547054322876, | |
| "eval_loss": 0.145263671875, | |
| "eval_runtime": 992.9783, | |
| "eval_samples_per_second": 1.126, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 3.2635854592980795, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.8255547054322876, | |
| "step": 5000, | |
| "total_flos": 1.0871892140465376e+21, | |
| "train_loss": 0.9307616455078125, | |
| "train_runtime": 97022.7442, | |
| "train_samples_per_second": 3.298, | |
| "train_steps_per_second": 0.052 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0871892140465376e+21, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |