GitHub - bfloat16/latent-diffusion-speech

Introduction

Personally modified version, for experimental testing only. >>> Original Branch Here <<<

The code is being cleaned, training and inference are currently unavailable, and keams will also be removed.

Collecting Japanese data and writing text cleaner.

Synced commits: 87b42cbaee19c90a914dbe4bfa5a527732739147

Unsynced commits:

Name		Name	Last commit message	Last commit date
Latest commit History 29 Commits
batch_proccessor		batch_proccessor
cluster		cluster
configs		configs
diffusion		diffusion
encoder		encoder
quantize		quantize
text		text
text2semantic		text2semantic
tools		tools
.gitignore		.gitignore
00_del_audio_over_30s.py		00_del_audio_over_30s.py
01_rename_to_increasing_number.py		01_rename_to_increasing_number.py
02_randomly_extract_valset.py		02_randomly_extract_valset.py
10_preprocess_train_unit.py		10_preprocess_train_unit.py
11_preprocess_train_mel.py		11_preprocess_train_mel.py
15_preprocess_train_text.py		15_preprocess_train_text.py
16_preprocess_train_tts.py		16_preprocess_train_tts.py
17_preprocess_train_cluster.py		17_preprocess_train_cluster.py
18_preprocess_val.py		18_preprocess_val.py
19_preprocess_token.py		19_preprocess_token.py
20_train_diffusion.py		20_train_diffusion.py
21_train_lm.py		21_train_lm.py
22_infer_tts.py		22_infer_tts.py
README.md		README.md
logo.png		logo.png
requirements.txt		requirements.txt