| { | |
| "best_metric": 1.9401395320892334, | |
| "best_model_checkpoint": "./zh-xlsr3/checkpoint-4000", | |
| "epoch": 12.158054711246201, | |
| "eval_steps": 2000, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 55.8891, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 9.8248, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 5.7785, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 5.5587, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 0.0003, | |
| "loss": 5.4992, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 0.00029811912225705323, | |
| "loss": 5.3469, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 0.00029623824451410655, | |
| "loss": 4.4282, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 0.00029435736677115986, | |
| "loss": 3.6204, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "learning_rate": 0.0002924764890282131, | |
| "loss": 3.2098, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "learning_rate": 0.00029059561128526643, | |
| "loss": 2.9893, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "eval_cer": 0.527875690283163, | |
| "eval_loss": 2.470449209213257, | |
| "eval_runtime": 144.8391, | |
| "eval_samples_per_second": 18.379, | |
| "eval_steps_per_second": 1.153, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "learning_rate": 0.00028871473354231974, | |
| "loss": 2.7826, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "learning_rate": 0.000286833855799373, | |
| "loss": 2.6828, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "learning_rate": 0.0002849529780564263, | |
| "loss": 2.5631, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.51, | |
| "learning_rate": 0.0002830721003134796, | |
| "loss": 2.4517, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 9.12, | |
| "learning_rate": 0.0002811912225705329, | |
| "loss": 2.3611, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.73, | |
| "learning_rate": 0.0002793103448275862, | |
| "loss": 2.2789, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 10.33, | |
| "learning_rate": 0.0002774294670846395, | |
| "loss": 2.1917, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 10.94, | |
| "learning_rate": 0.00027554858934169277, | |
| "loss": 2.124, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 11.55, | |
| "learning_rate": 0.00027366771159874603, | |
| "loss": 2.0215, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 12.16, | |
| "learning_rate": 0.00027178683385579934, | |
| "loss": 1.9656, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 12.16, | |
| "eval_cer": 0.41851721301844674, | |
| "eval_loss": 1.9401395320892334, | |
| "eval_runtime": 147.7922, | |
| "eval_samples_per_second": 18.012, | |
| "eval_steps_per_second": 1.13, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 32900, | |
| "num_train_epochs": 100, | |
| "save_steps": 2000, | |
| "total_flos": 2.4482031430853984e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |