227 lines
6.9 KiB
YAML
227 lines
6.9 KiB
YAML
name: export-kokoro-to-onnx
|
|
|
|
on:
|
|
push:
|
|
branches:
|
|
- export-kokoro
|
|
|
|
workflow_dispatch:
|
|
|
|
concurrency:
|
|
group: export-kokoro-to-onnx-${{ github.ref }}
|
|
cancel-in-progress: true
|
|
|
|
jobs:
|
|
export-kokoro-to-onnx:
|
|
if: github.repository_owner == 'k2-fsa' || github.repository_owner == 'csukuangfj'
|
|
name: export kokoro ${{ matrix.version }}
|
|
runs-on: ${{ matrix.os }}
|
|
strategy:
|
|
fail-fast: false
|
|
matrix:
|
|
os: [ubuntu-latest]
|
|
version: ["0.19", "1.0"]
|
|
python-version: ["3.10"]
|
|
|
|
steps:
|
|
- uses: actions/checkout@v4
|
|
|
|
- name: Setup Python ${{ matrix.python-version }}
|
|
uses: actions/setup-python@v5
|
|
with:
|
|
python-version: ${{ matrix.python-version }}
|
|
|
|
- name: Install Python dependencies
|
|
shell: bash
|
|
run: |
|
|
pip install "numpy<=1.26.4" onnx==1.16.0 onnxruntime==1.17.1 librosa soundfile piper_phonemize -f https://k2-fsa.github.io/icefall/piper_phonemize.html misaki[en] misaki[zh] torch==2.6.0+cpu -f https://download.pytorch.org/whl/torch
|
|
|
|
- name: Run
|
|
shell: bash
|
|
run: |
|
|
curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/espeak-ng-data.tar.bz2
|
|
tar xf espeak-ng-data.tar.bz2
|
|
rm espeak-ng-data.tar.bz2
|
|
cd scripts/kokoro
|
|
v=${{ matrix.version }}
|
|
if [[ $v = "0.19" ]]; then
|
|
./run.sh
|
|
elif [[ $v == "1.0" ]]; then
|
|
cd v1.0
|
|
./run.sh
|
|
fi
|
|
|
|
- name: Collect results ${{ matrix.version }}
|
|
if: matrix.version == '0.19'
|
|
shell: bash
|
|
run: |
|
|
src=scripts/kokoro
|
|
|
|
d=kokoro-en-v0_19
|
|
mkdir $d
|
|
cp -a LICENSE $d/LICENSE
|
|
cp -a espeak-ng-data $d/
|
|
cp -v $src/kokoro-v0_19.onnx $d/model.onnx
|
|
cp -v $src/voices.bin $d/
|
|
cp -v $src/tokens.txt $d/
|
|
cp -v $src/README-new.md $d/README.md
|
|
ls -lh $d/
|
|
tar cjfv $d.tar.bz2 $d
|
|
rm -rf $d
|
|
|
|
ls -lh $d.tar.bz2
|
|
|
|
- name: Collect results ${{ matrix.version }}
|
|
if: matrix.version == '1.0'
|
|
shell: bash
|
|
run: |
|
|
curl -SL -O https://github.com/csukuangfj/cppjieba/releases/download/sherpa-onnx-2024-04-19/dict.tar.bz2
|
|
tar xvf dict.tar.bz2
|
|
rm dict.tar.bz2
|
|
|
|
curl -SL -o date-zh.fst https://huggingface.co/csukuangfj/icefall-tts-aishell3-vits-low-2024-04-06/resolve/main/data/date.fst
|
|
curl -SL -o number-zh.fst https://huggingface.co/csukuangfj/icefall-tts-aishell3-vits-low-2024-04-06/resolve/main/data/number.fst
|
|
curl -SL -o phone-zh.fst https://huggingface.co/csukuangfj/icefall-tts-aishell3-vits-low-2024-04-06/resolve/main/data/phone.fst
|
|
|
|
src=scripts/kokoro/v1.0
|
|
|
|
d=kokoro-multi-lang-v1_0
|
|
mkdir $d
|
|
cp -a LICENSE $d/LICENSE
|
|
cp -a espeak-ng-data $d/
|
|
cp -v $src/kokoro.onnx $d/model.onnx
|
|
cp -v $src/voices.bin $d/
|
|
cp -v $src/tokens.txt $d/
|
|
cp -v $src/lexicon*.txt $d/
|
|
cp -v $src/README.md $d/README.md
|
|
cp -av dict $d/
|
|
cp -v ./*.fst $d/
|
|
ls -lh $d/
|
|
echo "---"
|
|
ls -lh $d/dict
|
|
|
|
tar cjfv $d.tar.bz2 $d
|
|
rm -rf $d
|
|
|
|
ls -lh $d.tar.bz2
|
|
|
|
- name: Publish to huggingface ${{ matrix.version }}
|
|
if: matrix.version == '0.19'
|
|
env:
|
|
HF_TOKEN: ${{ secrets.HF_TOKEN }}
|
|
uses: nick-fields/retry@v3
|
|
with:
|
|
max_attempts: 20
|
|
timeout_seconds: 200
|
|
shell: bash
|
|
command: |
|
|
git config --global user.email "csukuangfj@gmail.com"
|
|
git config --global user.name "Fangjun Kuang"
|
|
|
|
rm -rf huggingface
|
|
export GIT_LFS_SKIP_SMUDGE=1
|
|
export GIT_CLONE_PROTECTION_ACTIVE=false
|
|
|
|
git clone https://csukuangfj:$HF_TOKEN@huggingface.co/csukuangfj/kokoro-en-v0_19 huggingface
|
|
cd huggingface
|
|
rm -rf ./*
|
|
git fetch
|
|
git pull
|
|
|
|
git lfs track "cmn_dict"
|
|
git lfs track "ru_dict"
|
|
git lfs track "*.wav"
|
|
|
|
cp -a ../espeak-ng-data ./
|
|
mkdir -p test_wavs
|
|
|
|
cp -v ../scripts/kokoro/kokoro-v0_19.onnx ./model.onnx
|
|
|
|
cp -v ../scripts/kokoro/kokoro-v0_19-*.wav ./test_wavs/
|
|
|
|
cp -v ../scripts/kokoro/tokens.txt .
|
|
cp -v ../scripts/kokoro/voices.bin .
|
|
cp -v ../scripts/kokoro/README-new.md ./README.md
|
|
cp -v ../LICENSE ./
|
|
|
|
git lfs track "*.onnx"
|
|
git add .
|
|
|
|
ls -lh
|
|
|
|
git status
|
|
|
|
git commit -m "add models"
|
|
git push https://csukuangfj:$HF_TOKEN@huggingface.co/csukuangfj/kokoro-en-v0_19 main || true
|
|
|
|
- name: Publish to huggingface ${{ matrix.version }}
|
|
if: matrix.version == '1.0'
|
|
env:
|
|
HF_TOKEN: ${{ secrets.HF_TOKEN }}
|
|
uses: nick-fields/retry@v3
|
|
with:
|
|
max_attempts: 20
|
|
timeout_seconds: 200
|
|
shell: bash
|
|
command: |
|
|
git config --global user.email "csukuangfj@gmail.com"
|
|
git config --global user.name "Fangjun Kuang"
|
|
|
|
rm -rf huggingface
|
|
export GIT_LFS_SKIP_SMUDGE=1
|
|
export GIT_CLONE_PROTECTION_ACTIVE=false
|
|
|
|
git clone https://csukuangfj:$HF_TOKEN@huggingface.co/csukuangfj/kokoro-multi-lang-v1_0 huggingface
|
|
cd huggingface
|
|
rm -rf ./*
|
|
git fetch
|
|
git pull
|
|
|
|
git lfs track "cmn_dict"
|
|
git lfs track "ru_dict"
|
|
git lfs track "*.wav"
|
|
git lfs track "lexicon*.txt"
|
|
|
|
cp -a ../espeak-ng-data ./
|
|
|
|
cp -v ../scripts/kokoro/v1.0/kokoro.onnx ./model.onnx
|
|
|
|
|
|
cp -v ../scripts/kokoro/v1.0/tokens.txt .
|
|
cp -v ../scripts/kokoro/v1.0/voices.bin .
|
|
cp -v ../scripts/kokoro/v1.0/lexicon*.txt .
|
|
cp -v ../scripts/kokoro/v1.0/README.md ./README.md
|
|
cp -v ../LICENSE ./
|
|
cp -av ../dict ./
|
|
cp -v ../*.fst ./
|
|
|
|
git lfs track "*.onnx"
|
|
git add .
|
|
|
|
ls -lh
|
|
|
|
git status
|
|
|
|
git commit -m "add models"
|
|
git push https://csukuangfj:$HF_TOKEN@huggingface.co/csukuangfj/kokoro-multi-lang-v1_0 main || true
|
|
|
|
- name: Release
|
|
if: github.repository_owner == 'csukuangfj'
|
|
uses: svenstaro/upload-release-action@v2
|
|
with:
|
|
file_glob: true
|
|
file: ./*.tar.bz2
|
|
overwrite: true
|
|
repo_name: k2-fsa/sherpa-onnx
|
|
repo_token: ${{ secrets.UPLOAD_GH_SHERPA_ONNX_TOKEN }}
|
|
tag: tts-models
|
|
|
|
- name: Release
|
|
if: github.repository_owner == 'k2-fsa'
|
|
uses: svenstaro/upload-release-action@v2
|
|
with:
|
|
file_glob: true
|
|
file: ./*.tar.bz2
|
|
overwrite: true
|
|
tag: tts-models
|