| { |
| "kind": "pc", |
| "data": { |
| "sample_rate": 44100, |
| "hop_length": 512, |
| "n_fft": 2048, |
| "win_length": 2048, |
| "n_mels": 128, |
| "mel_fmin": 40.0, |
| "mel_fmax": 16000.0 |
| }, |
| "model": { |
| "sample_rate": 44100, |
| "num_mels": 128, |
| "upsample_initial_channel": 512, |
| "upsample_rates": [ |
| 8, |
| 8, |
| 2, |
| 2, |
| 2 |
| ], |
| "upsample_kernel_sizes": [ |
| 16, |
| 16, |
| 4, |
| 4, |
| 4 |
| ], |
| "resblock": "1", |
| "resblock_kernel_sizes": [ |
| 3, |
| 7, |
| 11 |
| ], |
| "resblock_dilation_sizes": [ |
| [ |
| 1, |
| 3, |
| 5 |
| ], |
| [ |
| 1, |
| 3, |
| 5 |
| ], |
| [ |
| 1, |
| 3, |
| 5 |
| ] |
| ], |
| "mini_nsf": true, |
| "harmonic_num": 8, |
| "noise_sigma": 0.0 |
| }, |
| "train": { |
| "batch_size": 12, |
| "crop_mel_frames": 50, |
| "learning_rate": 0.0001, |
| "finetune_learning_rate": 1e-05, |
| "betas": [ |
| 0.8, |
| 0.99 |
| ], |
| "weight_decay": 0.0, |
| "grad_clip": null, |
| "mel_loss_weight": 45.0, |
| "pc_wav_loss_weight": 30.0, |
| "pc_aug_rate": 0.4, |
| "pc_aug_key": 12.0, |
| "volume_aug_prob": 0.0, |
| "key_aug": false, |
| "key_aug_prob": 0.5, |
| "key_aug_min": 0.9, |
| "key_aug_max": 1.4, |
| "discriminator_warmup": 0, |
| "max_updates": 100000, |
| "checkpoint_interval": 2000, |
| "eval_interval": 2000, |
| "log_interval": 100, |
| "num_ckpt_keep": 5, |
| "num_valid_plots": 100, |
| "num_workers": 4, |
| "prefetch_factor": 2 |
| }, |
| "pretrained": "pc_nsf_hifigan_2025_02" |
| } |
|
|
