fish-speech/requirements.txt

# This file was autogenerated by uv via the following command:
#    uv pip compile pyproject.toml -o requirements.txt
absl-py==2.4.0
    # via tensorboard
aiofiles==24.1.0
    # via gradio
aiohappyeyeballs==2.6.1
    # via aiohttp
aiohttp==3.13.5
    # via
    #   datasets
    #   fsspec
aiosignal==1.4.0
    # via aiohttp
annotated-doc==0.0.4
    # via
    #   fastapi
    #   typer
annotated-types==0.7.0
    # via pydantic
antlr4-python3-runtime==4.9.3
    # via
    #   hydra-core
    #   omegaconf
anyio==4.13.0
    # via
    #   gradio
    #   httpx
    #   starlette
argbind==0.3.9
    # via
    #   descript-audio-codec
    #   descript-audiotools
asttokens==3.0.1
    # via stack-data
async-timeout==5.0.1
    # via aiohttp
attrs==26.1.0
    # via aiohttp
audioread==3.1.0
    # via librosa
baize==0.23.1
    # via kui
brotli==1.2.0
    # via gradio
cachetools==7.0.5
    # via fish-speech (pyproject.toml)
certifi==2026.2.25
    # via
    #   httpcore
    #   httpx
    #   requests
    #   sentry-sdk
cffi==2.0.0
    # via soundfile
charset-normalizer==3.4.7
    # via requests
click==8.3.2
    # via
    #   typer
    #   uvicorn
    #   wandb
contourpy==1.3.2
    # via matplotlib
cycler==0.12.1
    # via matplotlib
datasets==2.18.0
    # via fish-speech (pyproject.toml)
decorator==5.2.1
    # via
    #   ipython
    #   librosa
descript-audio-codec==1.0.0
    # via fish-speech (pyproject.toml)
descript-audiotools==0.7.2
    # via descript-audio-codec
dill==0.3.8
    # via
    #   datasets
    #   multiprocess
docstring-parser==0.17.0
    # via argbind
einops==0.8.2
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audio-codec
einx==0.2.2
    # via fish-speech (pyproject.toml)
exceptiongroup==1.3.1
    # via
    #   anyio
    #   ipython
executing==2.2.1
    # via stack-data
fastapi==0.135.3
    # via gradio
ffmpy==1.0.0
    # via
    #   descript-audiotools
    #   gradio
filelock==3.25.2
    # via
    #   datasets
    #   huggingface-hub
    #   torch
    #   transformers
fire==0.7.1
    # via randomname
flatten-dict==0.4.2
    # via descript-audiotools
fonttools==4.62.1
    # via matplotlib
frozendict==2.4.7
    # via einx
frozenlist==1.8.0
    # via
    #   aiohttp
    #   aiosignal
fsspec==2024.2.0
    # via
    #   datasets
    #   gradio-client
    #   huggingface-hub
    #   lightning
    #   pytorch-lightning
    #   torch
gitdb==4.0.12
    # via gitpython
gitpython==3.1.46
    # via wandb
gradio==6.11.0
    # via fish-speech (pyproject.toml)
gradio-client==2.4.0
    # via
    #   gradio
    #   hf-gradio
groovy==0.1.2
    # via gradio
grpcio==1.80.0
    # via
    #   fish-speech (pyproject.toml)
    #   tensorboard
h11==0.16.0
    # via
    #   httpcore
    #   uvicorn
hf-gradio==0.3.0
    # via gradio
hf-xet==1.4.3
    # via huggingface-hub
httpcore==1.0.9
    # via httpx
httpx==0.28.1
    # via
    #   gradio
    #   gradio-client
    #   safehttpx
huggingface-hub==0.36.2
    # via
    #   datasets
    #   gradio
    #   gradio-client
    #   tokenizers
    #   transformers
hydra-core==1.3.2
    # via fish-speech (pyproject.toml)
idna==3.11
    # via
    #   anyio
    #   httpx
    #   requests
    #   yarl
importlib-resources==6.5.2
    # via descript-audiotools
ipython==8.39.0
    # via descript-audiotools
jedi==0.19.2
    # via ipython
jinja2==3.1.6
    # via
    #   gradio
    #   torch
joblib==1.5.3
    # via
    #   librosa
    #   scikit-learn
julius==0.2.7
    # via descript-audiotools
kiwisolver==1.5.0
    # via matplotlib
kui==1.14.1
    # via fish-speech (pyproject.toml)
lazy-loader==0.5
    # via librosa
librosa==0.11.0
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audiotools
lightning==2.6.1
    # via fish-speech (pyproject.toml)
lightning-utilities==0.15.3
    # via
    #   lightning
    #   pytorch-lightning
    #   torchmetrics
llvmlite==0.47.0
    # via numba
loguru==0.7.3
    # via fish-speech (pyproject.toml)
loralib==0.1.2
    # via fish-speech (pyproject.toml)
markdown==3.10.2
    # via tensorboard
markdown-it-py==4.0.0
    # via rich
markdown2==2.5.5
    # via descript-audiotools
markupsafe==3.0.3
    # via
    #   gradio
    #   jinja2
    #   werkzeug
matplotlib==3.10.8
    # via descript-audiotools
matplotlib-inline==0.2.1
    # via ipython
mdurl==0.1.2
    # via markdown-it-py
modelscope==1.17.1
    # via fish-speech (pyproject.toml)
mpmath==1.3.0
    # via sympy
msgpack==1.1.2
    # via librosa
multidict==6.7.1
    # via
    #   aiohttp
    #   yarl
multiprocess==0.70.16
    # via datasets
natsort==8.4.0
    # via fish-speech (pyproject.toml)
networkx==3.4.2
    # via torch
numba==0.65.0
    # via
    #   librosa
    #   resampy
numpy==2.2.6
    # via
    #   fish-speech (pyproject.toml)
    #   contourpy
    #   datasets
    #   descript-audio-codec
    #   descript-audiotools
    #   einx
    #   gradio
    #   librosa
    #   matplotlib
    #   numba
    #   pandas
    #   pyloudnorm
    #   pystoi
    #   resampy
    #   scikit-learn
    #   scipy
    #   soundfile
    #   soxr
    #   tensorboard
    #   torch-stoi
    #   torchmetrics
    #   transformers
omegaconf==2.3.0
    # via hydra-core
opencc-python-reimplemented==0.1.7
    # via fish-speech (pyproject.toml)
orjson==3.11.8
    # via gradio
ormsgpack==1.12.2
    # via fish-speech (pyproject.toml)
packaging==26.0
    # via
    #   datasets
    #   gradio
    #   gradio-client
    #   huggingface-hub
    #   hydra-core
    #   lazy-loader
    #   lightning
    #   lightning-utilities
    #   matplotlib
    #   pooch
    #   pytorch-lightning
    #   silero-vad
    #   tensorboard
    #   torchmetrics
    #   transformers
    #   wandb
pandas==2.3.3
    # via
    #   datasets
    #   gradio
parso==0.8.6
    # via jedi
pexpect==4.9.0
    # via ipython
pillow==12.2.0
    # via
    #   gradio
    #   matplotlib
    #   tensorboard
platformdirs==4.9.6
    # via
    #   pooch
    #   wandb
pooch==1.9.0
    # via librosa
prompt-toolkit==3.0.52
    # via ipython
propcache==0.4.1
    # via
    #   aiohttp
    #   yarl
protobuf==5.29.6
    # via
    #   --override (workspace)
    #   descript-audiotools
    #   tensorboard
    #   wandb
ptyprocess==0.7.0
    # via pexpect
pure-eval==0.2.3
    # via stack-data
pyarrow==23.0.1
    # via datasets
pyarrow-hotfix==0.7
    # via datasets
pyaudio==0.2.14
    # via fish-speech (pyproject.toml)
pycparser==3.0
    # via cffi
pydantic==2.9.2
    # via
    #   fish-speech (pyproject.toml)
    #   fastapi
    #   gradio
    #   kui
    #   wandb
pydantic-core==2.23.4
    # via pydantic
pydub==0.25.1
    # via
    #   fish-speech (pyproject.toml)
    #   gradio
pygments==2.20.0
    # via
    #   ipython
    #   rich
pyloudnorm==0.2.0
    # via descript-audiotools
pyparsing==3.3.2
    # via matplotlib
pyrootutils==1.0.4
    # via fish-speech (pyproject.toml)
pystoi==0.4.1
    # via
    #   descript-audiotools
    #   torch-stoi
python-dateutil==2.9.0.post0
    # via
    #   matplotlib
    #   pandas
python-dotenv==1.2.2
    # via pyrootutils
python-multipart==0.0.24
    # via gradio
pytorch-lightning==2.6.1
    # via lightning
pytz==2026.1.post1
    # via
    #   gradio
    #   pandas
pyyaml==6.0.3
    # via
    #   argbind
    #   datasets
    #   gradio
    #   huggingface-hub
    #   lightning
    #   omegaconf
    #   pytorch-lightning
    #   transformers
    #   wandb
randomname==0.2.1
    # via descript-audiotools
regex==2026.4.4
    # via
    #   tiktoken
    #   transformers
requests==2.33.1
    # via
    #   datasets
    #   huggingface-hub
    #   modelscope
    #   pooch
    #   tiktoken
    #   transformers
    #   wandb
resampy==0.4.3
    # via fish-speech (pyproject.toml)
rich==14.3.3
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audiotools
    #   typer
safehttpx==0.1.7
    # via gradio
safetensors==0.7.0
    # via
    #   fish-speech (pyproject.toml)
    #   transformers
scikit-learn==1.7.2
    # via librosa
scipy==1.15.3
    # via
    #   descript-audiotools
    #   librosa
    #   pyloudnorm
    #   pystoi
    #   scikit-learn
semantic-version==2.10.0
    # via gradio
sentry-sdk==2.57.0
    # via wandb
setuptools==82.0.1
    # via tensorboard
shellingham==1.5.4
    # via typer
silero-vad==6.2.1
    # via fish-speech (pyproject.toml)
six==1.17.0
    # via
    #   flatten-dict
    #   python-dateutil
smmap==5.0.3
    # via gitdb
soundfile==0.13.1
    # via
    #   descript-audiotools
    #   librosa
soxr==1.0.0
    # via librosa
stack-data==0.6.3
    # via ipython
starlette==1.0.0
    # via
    #   fastapi
    #   gradio
sympy==1.14.0
    # via
    #   einx
    #   torch
tensorboard==2.20.0
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audiotools
tensorboard-data-server==0.7.2
    # via tensorboard
termcolor==3.3.0
    # via fire
threadpoolctl==3.6.0
    # via scikit-learn
tiktoken==0.12.0
    # via fish-speech (pyproject.toml)
tokenizers==0.22.2
    # via transformers
tomlkit==0.13.3
    # via gradio
torch==2.8.0
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audio-codec
    #   descript-audiotools
    #   einx
    #   julius
    #   lightning
    #   pytorch-lightning
    #   silero-vad
    #   torch-stoi
    #   torchaudio
    #   torchmetrics
torch-stoi==0.2.3
    # via descript-audiotools
torchaudio==2.8.0
    # via
    #   fish-speech (pyproject.toml)
    #   descript-audio-codec
    #   descript-audiotools
    #   silero-vad
    #   torch-stoi
torchmetrics==1.9.0
    # via
    #   lightning
    #   pytorch-lightning
tqdm==4.67.3
    # via
    #   datasets
    #   descript-audio-codec
    #   descript-audiotools
    #   huggingface-hub
    #   lightning
    #   modelscope
    #   pytorch-lightning
    #   transformers
traitlets==5.14.3
    # via
    #   ipython
    #   matplotlib-inline
transformers==4.57.3
    # via fish-speech (pyproject.toml)
typer==0.24.1
    # via
    #   gradio
    #   hf-gradio
typing-extensions==4.15.0
    # via
    #   aiosignal
    #   anyio
    #   exceptiongroup
    #   fastapi
    #   gradio
    #   gradio-client
    #   grpcio
    #   huggingface-hub
    #   ipython
    #   kui
    #   librosa
    #   lightning
    #   lightning-utilities
    #   multidict
    #   pydantic
    #   pydantic-core
    #   pytorch-lightning
    #   starlette
    #   torch
    #   typing-inspection
    #   uvicorn
    #   wandb
typing-inspection==0.4.2
    # via fastapi
tzdata==2026.1
    # via pandas
urllib3==2.6.3
    # via
    #   modelscope
    #   requests
    #   sentry-sdk
uvicorn==0.44.0
    # via
    #   fish-speech (pyproject.toml)
    #   gradio
wandb==0.25.1
    # via fish-speech (pyproject.toml)
wcwidth==0.6.0
    # via prompt-toolkit
werkzeug==3.1.8
    # via tensorboard
xxhash==3.6.0
    # via datasets
yarl==1.23.0
    # via aiohttp
zstandard==0.25.0
    # via fish-speech (pyproject.toml)