| { | |
| "best_metric": 18.683812405446293, | |
| "best_model_checkpoint": "./checkpoint-1000", | |
| "epoch": 42.10526315789474, | |
| "eval_steps": 1000, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 115.67640686035156, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 6.5238, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 27.986173629760742, | |
| "learning_rate": 9.800000000000001e-07, | |
| "loss": 4.874, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 15.896848678588867, | |
| "learning_rate": 1.48e-06, | |
| "loss": 3.0956, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 13.427591323852539, | |
| "learning_rate": 1.98e-06, | |
| "loss": 2.0652, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 13.170661926269531, | |
| "learning_rate": 2.4800000000000004e-06, | |
| "loss": 1.5245, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 6.32, | |
| "grad_norm": 12.342823028564453, | |
| "learning_rate": 2.9800000000000003e-06, | |
| "loss": 1.1716, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 12.604913711547852, | |
| "learning_rate": 3.46e-06, | |
| "loss": 0.8993, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "grad_norm": 11.273714065551758, | |
| "learning_rate": 3.96e-06, | |
| "loss": 0.6749, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "grad_norm": 10.14577579498291, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 0.4599, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 10.53, | |
| "grad_norm": 6.002610683441162, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 0.1846, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 11.58, | |
| "grad_norm": 2.2945444583892822, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 0.1001, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 12.63, | |
| "grad_norm": 2.423794984817505, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.0509, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 13.68, | |
| "grad_norm": 1.5324437618255615, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 0.0319, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 14.74, | |
| "grad_norm": 2.4710068702697754, | |
| "learning_rate": 6.96e-06, | |
| "loss": 0.0197, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 15.79, | |
| "grad_norm": 2.177290916442871, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.016, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 16.84, | |
| "grad_norm": 1.212317705154419, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.0111, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 17.89, | |
| "grad_norm": 1.4730677604675293, | |
| "learning_rate": 8.46e-06, | |
| "loss": 0.0093, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 18.95, | |
| "grad_norm": 0.6026259064674377, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.0062, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.0059149265289307, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.0072, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "grad_norm": 1.2911185026168823, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.0066, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 22.11, | |
| "grad_norm": 1.1619590520858765, | |
| "learning_rate": 9.94888888888889e-06, | |
| "loss": 0.0062, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 23.16, | |
| "grad_norm": 1.0487051010131836, | |
| "learning_rate": 9.893333333333334e-06, | |
| "loss": 0.0039, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 24.21, | |
| "grad_norm": 0.25514963269233704, | |
| "learning_rate": 9.837777777777778e-06, | |
| "loss": 0.0031, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 25.26, | |
| "grad_norm": 0.05470306798815727, | |
| "learning_rate": 9.782222222222222e-06, | |
| "loss": 0.0021, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 26.32, | |
| "grad_norm": 0.060789547860622406, | |
| "learning_rate": 9.726666666666668e-06, | |
| "loss": 0.0018, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 27.37, | |
| "grad_norm": 0.03411395475268364, | |
| "learning_rate": 9.671111111111112e-06, | |
| "loss": 0.0017, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 28.42, | |
| "grad_norm": 0.019400257617235184, | |
| "learning_rate": 9.615555555555558e-06, | |
| "loss": 0.0009, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 29.47, | |
| "grad_norm": 0.015261244960129261, | |
| "learning_rate": 9.56e-06, | |
| "loss": 0.0006, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 30.53, | |
| "grad_norm": 0.01666310615837574, | |
| "learning_rate": 9.504444444444446e-06, | |
| "loss": 0.0007, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 31.58, | |
| "grad_norm": 0.37447166442871094, | |
| "learning_rate": 9.44888888888889e-06, | |
| "loss": 0.0007, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 32.63, | |
| "grad_norm": 0.013453792780637741, | |
| "learning_rate": 9.393333333333334e-06, | |
| "loss": 0.0006, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 33.68, | |
| "grad_norm": 0.011604133062064648, | |
| "learning_rate": 9.33777777777778e-06, | |
| "loss": 0.0005, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 34.74, | |
| "grad_norm": 0.010738051496446133, | |
| "learning_rate": 9.282222222222222e-06, | |
| "loss": 0.0005, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 35.79, | |
| "grad_norm": 0.010525302961468697, | |
| "learning_rate": 9.226666666666668e-06, | |
| "loss": 0.0005, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 36.84, | |
| "grad_norm": 0.009034625254571438, | |
| "learning_rate": 9.171111111111112e-06, | |
| "loss": 0.0005, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 37.89, | |
| "grad_norm": 0.009822655469179153, | |
| "learning_rate": 9.115555555555556e-06, | |
| "loss": 0.0004, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 38.95, | |
| "grad_norm": 0.008321026340126991, | |
| "learning_rate": 9.060000000000001e-06, | |
| "loss": 0.0004, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.008215431123971939, | |
| "learning_rate": 9.004444444444445e-06, | |
| "loss": 0.0004, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 41.05, | |
| "grad_norm": 0.007949727587401867, | |
| "learning_rate": 8.94888888888889e-06, | |
| "loss": 0.0004, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 42.11, | |
| "grad_norm": 0.007843004539608955, | |
| "learning_rate": 8.893333333333333e-06, | |
| "loss": 0.0003, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 42.11, | |
| "eval_loss": 0.29718565940856934, | |
| "eval_runtime": 131.8407, | |
| "eval_samples_per_second": 2.89, | |
| "eval_steps_per_second": 0.728, | |
| "eval_wer": 18.683812405446293, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 218, | |
| "save_steps": 1000, | |
| "total_flos": 1.846946562048e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |