args.json

{
    "base_config": "config/valle.json",
    "dataset": [
        "LibriTTS",
    ],
    "model": {
        "add_prenet": false,
        "decoder_dim": 1024,
        "nar_scale_factor": 1,
        "nhead": 16,
        "norm_first": true,
        "num_decoder_layers": 12,
        "num_quantizers": 8,
        "prefix_mode": 1,
        "prepend_bos": false,
        "scaling_xformers": false,
        "share_embedding": true,
    },
    "model_type": "VALLE",
    "preprocess": {
        "acoustic_token_dir": "acoutic_tokens",
        "acoustic_token_extractor": "Encodec",
        "align_mel_duration": false,
        "audio_dir": "audios",
        "bits": 8,
        "contentvec_dir": "contentvec",
        "data_augment": false,
        "dur_dir": "durs",
        "duration_dir": "duration",
        "emo2id": "emo2id.json",
        "energy_dir": "energys",
        "energy_extract_mode": "from_mel",
        "energy_norm": false,
        "energy_remove_outlier": false,
        "extract_acoustic_token": true,
        "extract_audio": false,
        "extract_contentvec_feature": false,
        "extract_duration": false,
        "extract_energy": false,
        "extract_label": false,
        "extract_linear_spec": false,
        "extract_mcep": false,
        "extract_mel": false,
        "extract_mert_feature": false,
        "extract_phoneme": true,
        "extract_pitch": false,
        "extract_uv": false,
        "extract_wenet_feature": false,
        "extract_whisper_feature": false,
        "file_lst": "file.lst",
        "fmax": 12000,
        "fmin": 0,
        "hop_size": 120,
        "lab_dir": "labs",
        "label_dir": "labels",
        "lexicon_path": "./text/lexicon/librispeech-lexicon.txt",
        "linear_dir": "linears",
        "max_duration": 14,
        "mcep_dir": "mcep",
        "mel_dir": "mels",
        "mel_extract_mode": "",
        "mel_min_max_norm": false,
        "min_duration": 0.5,
        "min_level_db": -115,
        "n_fft": 1024,
        "n_mel": 80,
        "num_silent_frames": 8,
        "phone_dir": "phones",
        "phone_energy_dir": "phone_energys",
        "phone_pitch_dir": "phone_pitches",
        "phone_seq_file": "phone_seq_file",
        "pitch_dir": "pitches",
        "pitch_extractor": "parselmouth",
        "pitch_norm": false,
        "pitch_remove_outlier": false,
        "raw_data": "raw_data",
        "ref_level_db": 20,
        "sample_rate": 24000,
        "sampling_rate": "24000",
        "spk2id": "spk2id.json",
        "symbols_dict": "symbols.dict",
        "text_extractor": "espeak",
        "train_file": "train.json",
        "trim_fft_size": 512,
        "trim_hop_size": 128,
        "trim_silence": false,
        "trim_top_db": 30,
        "trimmed_wav_dir": "trimmed_wavs",
        "use_acoustic_token": true,
        "use_audio": false,
        "use_dur": false,
        "use_emoid": false,
        "use_frame_duration": false,
        "use_frame_energy": false,
        "use_frame_pitch": false,
        "use_lab": false,
        "use_label": false,
        "use_linear": false,
        "use_log_scale_energy": false,
        "use_log_scale_pitch": false,
        "use_mel": false,
        "use_min_max_norm_mel": false,
        "use_one_hot": false,
        "use_phn_seq": false,
        "use_phone": true,
        "use_phone_duration": false,
        "use_phone_energy": false,
        "use_phone_pitch": false,
        "use_spkid": false,
        "use_text": false,
        "use_uv": false,
        "use_wav": false,
        "use_wenet": false,
        "utt2emo": "utt2emo",
        "utt2spk": "utt2spk",
        "uv_dir": "uvs",
        "valid_file": "valid.json",
        "wav_dir": "wavs",
        "wenet_dir": "wenet",
        "win_size": 480,
    },
    "supported_model_type": [
        "GANVocoder",
        "Fastspeech2",
        "DiffSVC",
        "Transformer",
        "EDM",
        "CD",
    ],
    "train": {
        "adamw": {
            "lr": 0.0004,
        },
        "base_lr": 0.000125,
        "batch_size": 5,
        "dataloader": {
            "num_worker": 32,
            "pin_memory": true,
        },
        "ddp": false,
        "gradient_accumulation_step": 1,
        "keep_checkpoint_max": 5,
        "keep_last": [
            3,
            -1,
        ],
        "max_epoch": 100,
        "max_sentences": 8,
        "max_steps": 1000000,
        "max_tokens": 3600,
        "multi_speaker_training": false,
        "optimizer": "AdamW",
        "optimizer_name": "AdamW",
        "random_seed": 10086,
        "reducelronplateau": {
            "factor": 0.8,
            "min_lr": 0.0001,
            "patience": 10,
        },
        "reset_interval": 200,
        "run_eval": [
            false,
            true,
        ],
        "sampler": {
            "drop_last": true,
            "holistic_shuffle": true,
        },
        "save_checkpoint_stride": [
            1,
            1,
        ],
        "save_checkpoints_steps": 10000,
        "save_summary_steps": 500,
        "scheduler": "ReduceLROnPlateau",
        "scheduler_name": "Cosine",
        "start_epoch": 1,
        "total_training_steps": 50000,
        "tracker": [
            "tensorboard",
        ],
        "train_stage": 0,
        "valid_interval": 1000,
        "warmup_steps": 16000,
    },
    "use_custom_dataset": false,
}