mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-09-06 23:54:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix style

Browse Source
This commit is contained in:
yaozengwei 2023-11-29 19:57:16 +08:00
parent 0030d1b766
commit 70343a81f4
6 changed files with 19 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.flake8
View File

@ -15,7 +15,7 @@ per-file-ignores =
egs/librispeech/ASR/zipformer_mmi/*.py: E501, E203 egs/librispeech/ASR/zipformer_mmi/*.py: E501, E203
egs/librispeech/ASR/zipformer/*.py: E501, E203 egs/librispeech/ASR/zipformer/*.py: E501, E203
egs/librispeech/ASR/RESULTS.md: E999, egs/librispeech/ASR/RESULTS.md: E999,
egs/ljspeech/TTS/vits/*.py: E501, E203
# invalid escape sequence (cause by tex formular), W605 # invalid escape sequence (cause by tex formular), W605
icefall/utils.py: E501, W605 icefall/utils.py: E501, W605

6
docs/source/recipes/TTS/ljspeech/vits.rst
View File

@ -103,4 +103,10 @@ You can test the exported ONNX model with:
--model-filename vits/exp/vits-epoch-1000.onnx \ --model-filename vits/exp/vits-epoch-1000.onnx \
--tokens data/tokens.txt --tokens data/tokens.txt
Download pretrained models
--------------------------
If you don't want to train from scratch, you can download the pretrained models
by visiting the following link:
- `<https://huggingface.co/Zengwei/icefall-tts-ljspeech-vits-2023-11-29>`_

8
egs/ljspeech/TTS/local/compute_spectrogram_ljspeech.py
View File

@ -29,7 +29,13 @@ import os
from pathlib import Path from pathlib import Path
import torch import torch
from lhotse import CutSet, Spectrogram, SpectrogramConfig, LilcomChunkyWriter, load_manifest from lhotse import (
CutSet,
LilcomChunkyWriter,
Spectrogram,
SpectrogramConfig,
load_manifest,
)
from lhotse.audio import RecordingSet from lhotse.audio import RecordingSet
from lhotse.supervision import SupervisionSet from lhotse.supervision import SupervisionSet

2
egs/ljspeech/TTS/local/prepare_token_file.py
View File

@ -73,7 +73,7 @@ def get_token2id(manifest_file: Path) -> Dict[str, int]:
extra_tokens = [ extra_tokens = [
"<blk>", # 0 for blank "<blk>", # 0 for blank
"<sos/eos>", # 1 for sos and eos symbols. "<sos/eos>", # 1 for sos and eos symbols.
"<unk>" # 2 for OOV "<unk>", # 2 for OOV
] ]
all_tokens = set() all_tokens = set()

8
egs/ljspeech/TTS/local/prepare_tokens_ljspeech.py
View File

@ -34,9 +34,7 @@ def prepare_tokens_ljspeech():
suffix = "jsonl.gz" suffix = "jsonl.gz"
partition = "all" partition = "all"
cut_set = load_manifest( cut_set = load_manifest(output_dir / f"{prefix}_cuts_{partition}.{suffix}")
output_dir / f"{prefix}_cuts_{partition}.{suffix}"
)
g2p = g2p_en.G2p() g2p = g2p_en.G2p()
new_cuts = [] new_cuts = []
@ -51,9 +49,7 @@ def prepare_tokens_ljspeech():
new_cuts.append(cut) new_cuts.append(cut)
new_cut_set = CutSet.from_cuts(new_cuts) new_cut_set = CutSet.from_cuts(new_cuts)
new_cut_set.to_file( new_cut_set.to_file(output_dir / f"{prefix}_cuts_with_tokens_{partition}.{suffix}")
output_dir / f"{prefix}_cuts_with_tokens_{partition}.{suffix}"
)
if __name__ == "__main__": if __name__ == "__main__":

2
egs/ljspeech/TTS/vits/README.md
View File

@ -1 +1,3 @@
See https://k2-fsa.github.io/icefall/recipes/TTS/ljspeech/vits.html for detailed tutorials. See https://k2-fsa.github.io/icefall/recipes/TTS/ljspeech/vits.html for detailed tutorials.
Training logs, Tensorboard logs, and checkpoints are uploaded to https://huggingface.co/Zengwei/icefall-tts-ljspeech-vits-2023-11-29.