| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2916, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.00048, | |
| "loss": 7.0366, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 0.0004933240611961058, | |
| "loss": 0.4467, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.00048643949930458967, | |
| "loss": 0.3946, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.0004794853963838665, | |
| "loss": 0.399, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.0004726008344923505, | |
| "loss": 0.3696, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 0.3043254017829895, | |
| "eval_runtime": 65.1553, | |
| "eval_samples_per_second": 11.664, | |
| "eval_steps_per_second": 1.458, | |
| "eval_wer": 0.265993265993266, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.0004656467315716273, | |
| "loss": 0.3677, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.00045869262865090403, | |
| "loss": 0.3534, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 0.00045173852573018083, | |
| "loss": 0.34, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 0.0004447844228094576, | |
| "loss": 0.3428, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 0.0004378998609179416, | |
| "loss": 0.3533, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_loss": 0.27337372303009033, | |
| "eval_runtime": 65.7997, | |
| "eval_samples_per_second": 11.55, | |
| "eval_steps_per_second": 1.444, | |
| "eval_wer": 0.23931623931623933, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.00043094575799721834, | |
| "loss": 0.336, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 0.00042399165507649514, | |
| "loss": 0.3453, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 0.00041703755215577194, | |
| "loss": 0.342, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 0.0004100834492350487, | |
| "loss": 0.3232, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 0.0004031293463143255, | |
| "loss": 0.3419, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "eval_loss": 0.2687598764896393, | |
| "eval_runtime": 66.7716, | |
| "eval_samples_per_second": 11.382, | |
| "eval_steps_per_second": 1.423, | |
| "eval_wer": 0.23776223776223776, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 0.0003961752433936022, | |
| "loss": 0.3045, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 0.00038922114047287897, | |
| "loss": 0.327, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 0.0003822670375521558, | |
| "loss": 0.3162, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 0.00037531293463143256, | |
| "loss": 0.3111, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 0.0003684283727399166, | |
| "loss": 0.3399, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_loss": 0.2612040042877197, | |
| "eval_runtime": 66.305, | |
| "eval_samples_per_second": 11.462, | |
| "eval_steps_per_second": 1.433, | |
| "eval_wer": 0.22895622895622897, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 0.00036147426981919333, | |
| "loss": 0.3113, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 0.00035452016689847013, | |
| "loss": 0.307, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 0.0003475660639777469, | |
| "loss": 0.3061, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 0.0003406119610570236, | |
| "loss": 0.2974, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 0.00033365785813630047, | |
| "loss": 0.3087, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "eval_loss": 0.26245519518852234, | |
| "eval_runtime": 66.092, | |
| "eval_samples_per_second": 11.499, | |
| "eval_steps_per_second": 1.437, | |
| "eval_wer": 0.22986272986272988, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 0.0003267037552155772, | |
| "loss": 0.3007, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 0.00031974965229485396, | |
| "loss": 0.3073, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 0.00031279554937413076, | |
| "loss": 0.314, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 0.0003059109874826148, | |
| "loss": 0.2944, | |
| "step": 2900 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 7290, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2.020191562240091e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |