| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.603072983354673, | |
| "global_step": 45000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.9569123419101614e-05, | |
| "loss": 336.5875, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 1.869690361971217e-05, | |
| "loss": 219.0754, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 164.8076629638672, | |
| "eval_runtime": 187.3312, | |
| "eval_samples_per_second": 10.644, | |
| "eval_steps_per_second": 10.644, | |
| "eval_wer": 1.0029591701024072, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 1.7825119930222416e-05, | |
| "loss": 184.2771, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 1.6953336240732667e-05, | |
| "loss": 128.2162, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.6081116441343222e-05, | |
| "loss": 109.27, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_loss": 79.24954986572266, | |
| "eval_runtime": 187.7771, | |
| "eval_samples_per_second": 10.619, | |
| "eval_steps_per_second": 10.619, | |
| "eval_wer": 0.8314935496741588, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 1.5209332751853467e-05, | |
| "loss": 97.8815, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 1.4337112952464022e-05, | |
| "loss": 88.9743, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 61.599853515625, | |
| "eval_runtime": 188.2187, | |
| "eval_samples_per_second": 10.594, | |
| "eval_steps_per_second": 10.594, | |
| "eval_wer": 0.6520813938023673, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 1.3464893153074576e-05, | |
| "loss": 81.1935, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 1.2593109463584824e-05, | |
| "loss": 78.2627, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 1.1720889664195378e-05, | |
| "loss": 74.458, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "eval_loss": 54.358341217041016, | |
| "eval_runtime": 216.5604, | |
| "eval_samples_per_second": 9.208, | |
| "eval_steps_per_second": 9.208, | |
| "eval_wer": 0.5824577736401118, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 1.0849542084605322e-05, | |
| "loss": 71.4817, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "learning_rate": 9.977322285215875e-06, | |
| "loss": 66.9921, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "eval_loss": 49.08004379272461, | |
| "eval_runtime": 189.9086, | |
| "eval_samples_per_second": 10.5, | |
| "eval_steps_per_second": 10.5, | |
| "eval_wer": 0.5399321718313606, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "learning_rate": 9.105974705625819e-06, | |
| "loss": 66.8826, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 8.233754906236373e-06, | |
| "loss": 65.702, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 7.361535106846927e-06, | |
| "loss": 63.3176, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_loss": 47.168060302734375, | |
| "eval_runtime": 188.2915, | |
| "eval_samples_per_second": 10.59, | |
| "eval_steps_per_second": 10.59, | |
| "eval_wer": 0.5138648756483575, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 6.83, | |
| "learning_rate": 6.490187527256869e-06, | |
| "loss": 61.7128, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 7.26, | |
| "learning_rate": 5.6179677278674225e-06, | |
| "loss": 60.3547, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "eval_loss": 45.36619567871094, | |
| "eval_runtime": 217.9659, | |
| "eval_samples_per_second": 9.148, | |
| "eval_steps_per_second": 9.148, | |
| "eval_wer": 0.4957108658066232, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "learning_rate": 4.746184038377671e-06, | |
| "loss": 60.0223, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 8.11, | |
| "learning_rate": 3.8739642389882255e-06, | |
| "loss": 58.8691, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "learning_rate": 3.0017444395987794e-06, | |
| "loss": 58.391, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "eval_loss": 43.8188591003418, | |
| "eval_runtime": 187.5787, | |
| "eval_samples_per_second": 10.63, | |
| "eval_steps_per_second": 10.63, | |
| "eval_wer": 0.47882032185131, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "learning_rate": 2.1303968600087223e-06, | |
| "loss": 57.0914, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "learning_rate": 1.258177060619276e-06, | |
| "loss": 58.6047, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "eval_loss": 43.90667724609375, | |
| "eval_runtime": 188.2012, | |
| "eval_samples_per_second": 10.595, | |
| "eval_steps_per_second": 10.595, | |
| "eval_wer": 0.4761271445670967, | |
| "step": 45000 | |
| } | |
| ], | |
| "max_steps": 46860, | |
| "num_train_epochs": 10, | |
| "total_flos": 5.571669557707454e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |