| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 30.0, | |
| "global_step": 27510, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.982e-05, | |
| "loss": 4.6277, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2.9449092928544985e-05, | |
| "loss": 1.5086, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2.889374305812662e-05, | |
| "loss": 0.7041, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2.834061458718993e-05, | |
| "loss": 0.5814, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2.7785264716771567e-05, | |
| "loss": 0.5086, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 2.7232136245834875e-05, | |
| "loss": 0.462, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 2.667789707515735e-05, | |
| "loss": 0.4247, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "learning_rate": 2.6123657904479824e-05, | |
| "loss": 0.3919, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "learning_rate": 2.556830803406146e-05, | |
| "loss": 0.3743, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "learning_rate": 2.5014068863383933e-05, | |
| "loss": 0.3452, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "eval_loss": 0.38393500447273254, | |
| "eval_runtime": 207.46, | |
| "eval_samples_per_second": 59.009, | |
| "eval_steps_per_second": 1.846, | |
| "eval_wer": 0.4573501807782915, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 2.445871899296557e-05, | |
| "loss": 0.3357, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 2.3904479822288042e-05, | |
| "loss": 0.3131, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "learning_rate": 2.334912995186968e-05, | |
| "loss": 0.3039, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.63, | |
| "learning_rate": 2.279489078119215e-05, | |
| "loss": 0.2879, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "learning_rate": 2.2240651610514624e-05, | |
| "loss": 0.2772, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "learning_rate": 2.1686412439837097e-05, | |
| "loss": 0.2649, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.27, | |
| "learning_rate": 2.1132173269159573e-05, | |
| "loss": 0.2596, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 9.81, | |
| "learning_rate": 2.057682339874121e-05, | |
| "loss": 0.2526, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "learning_rate": 2.0021473528322843e-05, | |
| "loss": 0.2451, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 10.91, | |
| "learning_rate": 1.946834505738615e-05, | |
| "loss": 0.2466, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.91, | |
| "eval_loss": 0.40105822682380676, | |
| "eval_runtime": 206.4585, | |
| "eval_samples_per_second": 59.295, | |
| "eval_steps_per_second": 1.855, | |
| "eval_wer": 0.4067448569520906, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 11.45, | |
| "learning_rate": 1.8914105886708627e-05, | |
| "loss": 0.2564, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 1.8358756016290264e-05, | |
| "loss": 0.3219, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 12.54, | |
| "learning_rate": 1.7804516845612736e-05, | |
| "loss": 0.3501, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 13.09, | |
| "learning_rate": 1.7249166975194373e-05, | |
| "loss": 0.375, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 13.63, | |
| "learning_rate": 1.669381710477601e-05, | |
| "loss": 0.3621, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 14.18, | |
| "learning_rate": 1.614068863383932e-05, | |
| "loss": 0.4168, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 14.72, | |
| "learning_rate": 1.5585338763420955e-05, | |
| "loss": 0.515, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 15.27, | |
| "learning_rate": 1.5033320992225103e-05, | |
| "loss": 0.8026, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 15.81, | |
| "learning_rate": 1.4477971121806738e-05, | |
| "loss": 1.3623, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 16.36, | |
| "learning_rate": 1.3922621251388376e-05, | |
| "loss": 1.5753, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 16.36, | |
| "eval_loss": 1.2936806678771973, | |
| "eval_runtime": 205.5873, | |
| "eval_samples_per_second": 59.546, | |
| "eval_steps_per_second": 1.863, | |
| "eval_wer": 0.8843522846092795, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 16.9, | |
| "learning_rate": 1.3368382080710849e-05, | |
| "loss": 1.6596, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 17.45, | |
| "learning_rate": 1.2813032210292485e-05, | |
| "loss": 1.6553, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 17.99, | |
| "learning_rate": 1.2258793039614958e-05, | |
| "loss": 1.7339, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 18.54, | |
| "learning_rate": 1.1703443169196594e-05, | |
| "loss": 1.8133, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 19.08, | |
| "learning_rate": 1.1149203998519067e-05, | |
| "loss": 1.7555, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "learning_rate": 1.059496482784154e-05, | |
| "loss": 1.7767, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 20.17, | |
| "learning_rate": 1.0039614957423176e-05, | |
| "loss": 1.8686, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 20.72, | |
| "learning_rate": 9.485375786745649e-06, | |
| "loss": 1.8748, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 21.26, | |
| "learning_rate": 8.931136616068122e-06, | |
| "loss": 1.9346, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 21.81, | |
| "learning_rate": 8.37578674564976e-06, | |
| "loss": 1.9454, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 21.81, | |
| "eval_loss": 1.822685718536377, | |
| "eval_runtime": 205.1767, | |
| "eval_samples_per_second": 59.666, | |
| "eval_steps_per_second": 1.867, | |
| "eval_wer": 0.9391778682019174, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 22.36, | |
| "learning_rate": 7.820436875231396e-06, | |
| "loss": 1.9335, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 22.9, | |
| "learning_rate": 7.267308404294706e-06, | |
| "loss": 1.9261, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 23.45, | |
| "learning_rate": 6.7130692336171785e-06, | |
| "loss": 1.9216, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 23.99, | |
| "learning_rate": 6.158830062939652e-06, | |
| "loss": 1.9228, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 24.54, | |
| "learning_rate": 5.604590892262125e-06, | |
| "loss": 1.9233, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 25.08, | |
| "learning_rate": 5.049241021843762e-06, | |
| "loss": 1.9197, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 25.63, | |
| "learning_rate": 4.495001851166235e-06, | |
| "loss": 1.9206, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 26.17, | |
| "learning_rate": 3.939651980747871e-06, | |
| "loss": 1.9235, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 26.72, | |
| "learning_rate": 3.384302110329508e-06, | |
| "loss": 1.9229, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 27.26, | |
| "learning_rate": 2.8311736393928178e-06, | |
| "loss": 1.922, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 27.26, | |
| "eval_loss": 1.8085108995437622, | |
| "eval_runtime": 206.76, | |
| "eval_samples_per_second": 59.209, | |
| "eval_steps_per_second": 1.852, | |
| "eval_wer": 0.9420627621915116, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 27.81, | |
| "learning_rate": 2.275823768974454e-06, | |
| "loss": 1.9211, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 28.35, | |
| "learning_rate": 1.7204738985560903e-06, | |
| "loss": 1.9205, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 28.9, | |
| "learning_rate": 1.1662347278785636e-06, | |
| "loss": 1.9253, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 29.44, | |
| "learning_rate": 6.119955572010366e-07, | |
| "loss": 1.92, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 29.99, | |
| "learning_rate": 5.6645686782673083e-08, | |
| "loss": 1.9231, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "step": 27510, | |
| "total_flos": 6.935706045316176e+19, | |
| "train_loss": 1.198725398988214, | |
| "train_runtime": 26753.1928, | |
| "train_samples_per_second": 65.77, | |
| "train_steps_per_second": 1.028 | |
| } | |
| ], | |
| "max_steps": 27510, | |
| "num_train_epochs": 30, | |
| "total_flos": 6.935706045316176e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |