Fine-tune with my own dataset, wer is 1

[abcdbosh]

finetune according to this blog： https://huggingface.co/blog/fine-tune-wav2vec2-english， But this problem arises，Here it is train_state.json: "best_metric": null, "best_model_checkpoint": null, "epoch": 18.9873417721519, "global_step": 1500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 6.33, "learning_rate": 3.49e-05, "loss": 14.8414, "step": 500 }, { "epoch": 6.33, "eval_loss": NaN, "eval_runtime": 36.5218, "eval_samples_per_second": 49.286, "eval_steps_per_second": 2.054, "eval_wer": 0.9999660072064722, "step": 500 }, { "epoch": 12.66, "learning_rate": 8.34e-05, "loss": 0.0, "step": 1000 }, { "epoch": 12.66, "eval_loss": NaN, "eval_runtime": 36.5778, "eval_samples_per_second": 49.21, "eval_steps_per_second": 2.05, "eval_wer": 0.9999660072064722, "step": 1000 }, { "epoch": 18.99, "learning_rate": 7.562043795620438e-05, "loss": 0.0, "step": 1500 }, { "epoch": 18.99, "eval_loss": NaN, "eval_runtime": 36.5273, "eval_samples_per_second": 49.278, "eval_steps_per_second": 2.053, "eval_wer": 0.9999660072064722, "step": 1500 } ], "max_steps": 2370, "num_train_epochs": 30, "total_flos": 4.074416087326559e+18, "trial_name": null, "trial_params": null }

What could be the problem?