| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2224, | |
| "eval_steps": 500, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 4.1002, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 3.96e-06, | |
| "loss": 3.4625, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 2.4907, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 1.5881, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 1.1583, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_combined_wer": 0.3216761184625079, | |
| "eval_f1_score": 0.07455306200076076, | |
| "eval_label_f1": 0.1414986686953214, | |
| "eval_loss": 1.0361335277557373, | |
| "eval_runtime": 337.7385, | |
| "eval_samples_per_second": 2.961, | |
| "eval_steps_per_second": 0.047, | |
| "eval_wer": 0.20672983948946044, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.98830238119205e-06, | |
| "loss": 0.9384, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.952307128483257e-06, | |
| "loss": 0.7396, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.892184733248666e-06, | |
| "loss": 0.4787, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.808228105754378e-06, | |
| "loss": 0.436, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.700846274250252e-06, | |
| "loss": 0.4069, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_combined_wer": 0.22034499054820417, | |
| "eval_f1_score": 0.4222737819025522, | |
| "eval_label_f1": 0.5939675174013921, | |
| "eval_loss": 0.4110757112503052, | |
| "eval_runtime": 259.1805, | |
| "eval_samples_per_second": 3.858, | |
| "eval_steps_per_second": 0.062, | |
| "eval_wer": 0.12349642235544382, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.570562392225395e-06, | |
| "loss": 0.4044, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.418011189656942e-06, | |
| "loss": 0.3882, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.24393588066941e-06, | |
| "loss": 0.3855, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.0491845426702e-06, | |
| "loss": 0.3774, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 8.834705984601708e-06, | |
| "loss": 0.3708, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_combined_wer": 0.2200693131695022, | |
| "eval_f1_score": 0.4609408718170047, | |
| "eval_label_f1": 0.6266724212343547, | |
| "eval_loss": 0.3768249750137329, | |
| "eval_runtime": 253.1584, | |
| "eval_samples_per_second": 3.95, | |
| "eval_steps_per_second": 0.063, | |
| "eval_wer": 0.12949139431444595, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 8.601545124439535e-06, | |
| "loss": 0.3629, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 8.350837898457142e-06, | |
| "loss": 0.3609, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 8.083805727058514e-06, | |
| "loss": 0.3605, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 7.801749564140724e-06, | |
| "loss": 0.3534, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 7.5060435589773215e-06, | |
| "loss": 0.3512, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_combined_wer": 0.22231411468178955, | |
| "eval_f1_score": 0.5141859450021825, | |
| "eval_label_f1": 0.6835443037974683, | |
| "eval_loss": 0.3623911738395691, | |
| "eval_runtime": 245.3233, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.065, | |
| "eval_wer": 0.13587313865789982, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 7.1981283615012e-06, | |
| "loss": 0.3504, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 6.879504103602934e-06, | |
| "loss": 0.3475, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 6.551723090639008e-06, | |
| "loss": 0.3433, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 6.216382238756147e-06, | |
| "loss": 0.3397, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5.8751152948763815e-06, | |
| "loss": 0.3411, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_combined_wer": 0.22038437303087588, | |
| "eval_f1_score": 0.5225073622212874, | |
| "eval_label_f1": 0.6882625157761885, | |
| "eval_loss": 0.3542915880680084, | |
| "eval_runtime": 266.2481, | |
| "eval_samples_per_second": 3.756, | |
| "eval_steps_per_second": 0.06, | |
| "eval_wer": 0.13738155095726165, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.52958487724626e-06, | |
| "loss": 0.337, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5.18147437532788e-06, | |
| "loss": 0.3346, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 4.832479748494643e-06, | |
| "loss": 0.3348, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.484301263487664e-06, | |
| "loss": 0.3316, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 4.138635210887117e-06, | |
| "loss": 0.3313, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_combined_wer": 0.22353497164461247, | |
| "eval_f1_score": 0.5193370165745856, | |
| "eval_label_f1": 0.6808329791755205, | |
| "eval_loss": 0.34923481941223145, | |
| "eval_runtime": 238.9408, | |
| "eval_samples_per_second": 4.185, | |
| "eval_steps_per_second": 0.067, | |
| "eval_wer": 0.1397795397408625, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 3.797165640955041e-06, | |
| "loss": 0.3322, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.4615561591117486e-06, | |
| "loss": 0.3258, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 3.1334418210174268e-06, | |
| "loss": 0.3297, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 2.814421166745337e-06, | |
| "loss": 0.328, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 2.506048432855247e-06, | |
| "loss": 0.3252, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_combined_wer": 0.22511027095148078, | |
| "eval_f1_score": 0.533276812208563, | |
| "eval_label_f1": 0.6892751165748198, | |
| "eval_loss": 0.34593451023101807, | |
| "eval_runtime": 242.7141, | |
| "eval_samples_per_second": 4.12, | |
| "eval_steps_per_second": 0.066, | |
| "eval_wer": 0.14356990910848966, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 2.209825980309151e-06, | |
| "loss": 0.3262, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 1.927196975119678e-06, | |
| "loss": 0.3247, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.6595383573903412e-06, | |
| "loss": 0.3246, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 1.4081541330017706e-06, | |
| "loss": 0.3197, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.1742690206261293e-06, | |
| "loss": 0.3293, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_combined_wer": 0.2236531190926276, | |
| "eval_f1_score": 0.5324947589098533, | |
| "eval_label_f1": 0.6859538784067086, | |
| "eval_loss": 0.34473055601119995, | |
| "eval_runtime": 244.5187, | |
| "eval_samples_per_second": 4.09, | |
| "eval_steps_per_second": 0.065, | |
| "eval_wer": 0.14159736994778574, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 1.4916379341570048e+20, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |