Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

configpy updates including TTSDataset cached mode

This commit is contained in:
Eren Golge 2018-12-17 16:37:39 +01:00
parent 8ff9253abd
commit c245cb4f75
1 changed files with 12 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
config.json
View File

@ -1,5 +1,5 @@
{ {
"model_name": "TTS-master", "model_name": "TTS-dev-tweb",
"model_description": "Higher dropout rate for stopnet and disabled custom initialization, pull current mel prediction to stopnet.", "model_description": "Higher dropout rate for stopnet and disabled custom initialization, pull current mel prediction to stopnet.",
"audio":{ "audio":{
@ -22,18 +22,18 @@
"clip_norm": true, // clip normalized values into the range. "clip_norm": true, // clip normalized values into the range.
"mel_fmin": null, // minimum freq level for mel-spec. ~50 for male and ~95 for female voices. Tune for dataset!! "mel_fmin": null, // minimum freq level for mel-spec. ~50 for male and ~95 for female voices. Tune for dataset!!
"mel_fmax": null, // maximum freq level for mel-spec. Tune for dataset!! "mel_fmax": null, // maximum freq level for mel-spec. Tune for dataset!!
"do_trim_silence": true // enable trimming of slience of audio as you load it. "do_trim_silence": false // enable trimming of slience of audio as you load it. LJspeech (false), TWEB (false), Nancy (true)
}, },
"embedding_size": 256, "embedding_size": 256,
"text_cleaner": "english_cleaners", "text_cleaner": "english_cleaners",
"epochs": 1000, "epochs": 1000,
"lr": 0.0001, "lr": 0.001,
"lr_decay": false, "lr_decay": false,
"warmup_steps": 4000, "warmup_steps": 4000,
"batch_size": 32, "batch_size": 20,
"eval_batch_size":32, "eval_batch_size":32,
"r": 5, "r": 5,
"wd": 0.000001, "wd": 0.000001,
@ -44,12 +44,12 @@
"run_eval": true, "run_eval": true,
"data_path": "../../Data/LJSpeech-1.1/", // can overwritten from command argument "data_path": "../../Data/LJSpeech-1.1/", // can overwritten from command argument
"meta_file_train": "prompts_train.data", // metafile for training dataloader "meta_file_train": "transcript.txt", // metafile for training dataloader.
"meta_file_val": "prompts_val.data", // metafile for validation dataloader "meta_file_val": "", // metafile for evaluation dataloader.
"data_loader": "TTSDataset", // dataloader, ["TTSDataset", "TTSDatasetCached", "TTSDatasetMemory"] "dataset": "tweb", // one of TTS.dataset.preprocessors depending on your target dataset. Use "tts_cache" for pre-computed dataset by extract_features.py
"dataset": "nancy", // one of TTS.dataset.preprocessors, only valid id dataloader == "TTSDataset", rest uses "tts_cache" by default. "min_seq_len": 0, // minimum text length to use in training
"min_seq_len": 0, "max_seq_len": 300, // maximum text length
"output_path": "../keep/", "output_path": "../keep/", // output path for all training outputs.
"num_loader_workers": 8, "num_loader_workers": 8, // number of training data loader processes. Don't set it too big. 4-8 are good values.
"num_val_loader_workers": 4 "num_val_loader_workers": 4 // number of evaluation data loader processes.
} }