| { | |
| "exp_name": "F5TTS_Base", | |
| "learning_rate": 1e-05, | |
| "batch_size_per_gpu": 300, | |
| "batch_size_type": "frame", | |
| "max_samples": 32, | |
| "grad_accumulation_steps": 1, | |
| "max_grad_norm": 1.0, | |
| "epochs": 10, | |
| "num_warmup_updates": 1000, | |
| "save_per_updates": 0, | |
| "last_per_steps": 0, | |
| "save_every_epochs": 5, | |
| "finetune": true, | |
| "file_checkpoint_train": "/workspace/F5-TTS/ckpts/firstpixelptbr/model_last.pt", | |
| "tokenizer_type": "custom", | |
| "tokenizer_file": "/workspace/F5-TTS/src/f5_tts/infer/examples/vocab.txt", | |
| "mixed_precision": "none", | |
| "logger": "wandb", | |
| "bnb_optimizer": false | |
| } |