49 lines
1.9 KiB
Python
49 lines
1.9 KiB
Python
#!/usr/bin/env python3
|
|
# -*- coding: utf-8 -*-
|
|
|
|
####################################################################################
|
|
### ###
|
|
### Configuration File for Multi-Speaker Baseline Model Training & Voice Cloning ###
|
|
### ###
|
|
####################################################################################
|
|
|
|
import os
|
|
|
|
# Data Sets
|
|
|
|
## VCTK (v0.92), sampling rate: 48000
|
|
|
|
VCTK_PRESET = "VCTK"
|
|
VCTK_DATASET_NAME = "VCTK"
|
|
VCTK_DATASET_FORMATTER = "vctk"
|
|
VCTK_DATASET_FILE_FORMAT = "flac"
|
|
VCTK_DATASET_PATH = "results/datasets/sr22050/VCTK-Corpus-0.92"
|
|
VCTK_SPK_EMB_PATH = os.path.join(VCTK_DATASET_PATH, "speakers.pth")
|
|
|
|
## LibriTTS TC360, sampling rate: 24000
|
|
|
|
LIBRITTS_TC360_PRESET = "LibriTTS_tc360"
|
|
LIBRITTS_TC360_DATASET_NAME = "LibtriTTS-tc360"
|
|
LIBRITTS_TC360_DATASET_FORMATTER = "libri_tts"
|
|
LIBRITTS_TC360_DATASET_FILE_FORMAT = "wav"
|
|
LIBRITTS_TC360_DATASET_PATH = "results/datasets/sr22050/LibriTTS/train-clean-360"
|
|
LIBRITTS_TC360_SPK_EMB_PATH = os.path.join(LIBRITTS_TC360_DATASET_PATH, "speakers.pth")
|
|
|
|
## DAPS
|
|
|
|
## Potion salutation recordings
|
|
|
|
POTION_SALUT_PRESET = "POTION_Salut"
|
|
POTION_SALUT_DATASET_NAME = "potion-Salut"
|
|
POTION_SALUT_DATASET_FORMATTER = "vctk_old"
|
|
POTION_SALUT_DATASET_FILE_FORMAT = "wav"
|
|
POTION_SALUT_DATASET_PATH = "results/datasets/sr22050/potion-salut-corpus-4ac24ce8-8405-4b70-8b48-018d4492f6e9"
|
|
POTION_SALUT_SPK_EMB_PATH = os.path.join(POTION_SALUT_DATASET_PATH, "speakers.pth")
|
|
|
|
## Potion voice cloning recordings
|
|
|
|
POTION_SALUT_PRESET = "potion_voice_cloning"
|
|
POTION_SALUT_DATASET_NAME = ""
|
|
POTION_SALUT_DATASET_FORMATTER = "vctk_old"
|
|
POTION_SALUT_DATASET_FILE_FORMAT = "wav"
|