| accelerator: gpu |
| accumulate_grad_batches: null |
| amp_backend: null |
| amp_level: null |
| auto_lr_find: false |
| auto_scale_batch_size: false |
| auto_select_gpus: null |
| batch_size: 96 |
| benchmark: null |
| betas: !!python/tuple |
| - 0.8 |
| - 0.99 |
| c_kl: 1.0 |
| c_mel: 45 |
| channels: 1 |
| check_val_every_n_epoch: 1 |
| checkpoint_epochs: 10 |
| dataset: |
| - !!python/object/apply:pathlib.PosixPath |
| - / |
| - data |
| - piper_css10_ja_preprocessed |
| - dataset.jsonl |
| dataset_dir: &id001 !!python/object/apply:pathlib.PosixPath |
| - / |
| - data |
| - piper_css10_ja_preprocessed |
| default_root_dir: *id001 |
| detect_anomaly: false |
| devices: '1' |
| enable_checkpointing: true |
| enable_model_summary: true |
| enable_progress_bar: true |
| eps: 1.0e-09 |
| fast_dev_run: false |
| filter_channels: 768 |
| filter_length: 1024 |
| gin_channels: 0 |
| gpus: null |
| grad_clip: null |
| gradient_clip_algorithm: null |
| gradient_clip_val: null |
| hidden_channels: 192 |
| hop_length: 256 |
| inference_mode: true |
| init_lr_ratio: 1.0 |
| inter_channels: 192 |
| ipus: null |
| kernel_size: 3 |
| learning_rate: 0.0002 |
| limit_predict_batches: null |
| limit_test_batches: null |
| limit_train_batches: null |
| limit_val_batches: null |
| log_every_n_steps: 50 |
| logger: true |
| lr_decay: 0.999875 |
| max_epochs: 100 |
| max_phoneme_ids: null |
| max_steps: -1 |
| max_time: null |
| mel_channels: 80 |
| mel_fmax: null |
| mel_fmin: 0.0 |
| min_epochs: null |
| min_steps: null |
| move_metrics_to_cpu: false |
| multiple_trainloader_mode: max_size_cycle |
| n_heads: 2 |
| n_layers: 6 |
| n_layers_q: 3 |
| num_nodes: 1 |
| num_processes: null |
| num_sanity_val_steps: 2 |
| num_speakers: 1 |
| num_symbols: 53 |
| num_test_examples: 5 |
| num_workers: 45 |
| overfit_batches: 0.0 |
| p_dropout: 0.1 |
| plugins: null |
| precision: 16 |
| profiler: null |
| quality: medium |
| reload_dataloaders_every_n_epochs: 0 |
| replace_sampler_ddp: true |
| resblock: '2' |
| resblock_dilation_sizes: !!python/tuple |
| - !!python/tuple |
| - 1 |
| - 2 |
| - !!python/tuple |
| - 2 |
| - 6 |
| - !!python/tuple |
| - 3 |
| - 12 |
| resblock_kernel_sizes: !!python/tuple |
| - 3 |
| - 5 |
| - 7 |
| resume_from_checkpoint: null |
| resume_from_single_speaker_checkpoint: null |
| sample_bytes: 2 |
| sample_rate: 22050 |
| save_top_k: -1 |
| seed: 1234 |
| segment_size: 8192 |
| strategy: null |
| sync_batchnorm: false |
| tpu_cores: null |
| track_grad_norm: -1 |
| upsample_initial_channel: 256 |
| upsample_kernel_sizes: !!python/tuple |
| - 16 |
| - 16 |
| - 8 |
| upsample_rates: !!python/tuple |
| - 8 |
| - 8 |
| - 4 |
| use_sdp: true |
| use_spectral_norm: false |
| val_check_interval: null |
| validation_split: 0.1 |
| warmup_epochs: 0 |
| win_length: 1024 |
|
|
