pyproject.toml

[build-system]
requires = ["setuptools>=61.0"]
build-backend = "setuptools.build_meta"

[project]
name = "fastvideo"
version = "0.2.0"
description = "FastVideo"
readme = "README.md"
requires-python = ">=3.10"
classifiers = [
    "Programming Language :: Python :: 3",
    "License :: OSI Approved :: Apache Software License",
]

dependencies = [
    # Core Libraries
    "scipy>=1.14.1",
    "six>=1.16.0",
    "h5py>=3.12.1",
    "requests>=2.32.2",

    # Machine Learning & Transformers
    # GLM-Image's AR encoder (GlmImageForConditionalGeneration) first ships in
    # transformers 5.0.0; floor bumped from >=4.57.3 to >=5.0.0 (stable, not rc).
    "transformers>=5.0.0",
    # <0.23: tokenizers 0.23 renamed RobertaProcessing's binding args, so
    # transformers' CLIP-style tokenizer loading dies with
    # "RobertaProcessing.__new__() got an unexpected keyword argument 'cls'".
    "tokenizers>=0.20.1,<0.23",
    "sentencepiece>=0.2.0",
    "timm>=1.0.11",
    "peft>=0.15.0",
    "diffusers>=0.38.0",
    "torch==2.12.0",
    "torchvision",
    "torchaudio",
    # CUDA-only: no wheels for non-Linux platforms, and the sdist build shells
    # out to nvcc unconditionally. Gate on sys_platform so installs on macOS
    # (MPS) and other non-Linux platforms don't try to build it from source.
    "flashinfer-python; sys_platform == 'linux'",

    # Acceleration & Optimization
    "accelerate==1.0.1",

    # Computer Vision & Image Processing
    "opencv-python>=4.10.0.84",
    "pillow>=10.3.0",
    "imageio>=2.36.0",
    "imageio-ffmpeg>=0.5.1",
    "einops",

    # Experiment Tracking & Logging
    "wandb>=0.21.0",
    "loguru",
    "test-tube>=0.7.5",
    
    # Miscellaneous Utilities
    "tqdm",
    "pytest",
    "PyYAML>=6.0.1",
    "protobuf>=5.28.3",
    "gradio==5.32.0",
    "moviepy>=2.0.0",
    "flask",
    "flask_restful",
    "aiohttp",
    "aiofiles",
    "huggingface_hub",
    "cloudpickle",
    "omegaconf",

    # System & Monitoring Tools
    "gpustat",
    "watch",
    "remote-pdb",

    # Kernel & Packaging
    # CUDA-only: PyPI only publishes manylinux wheels, and the sdist build
    # requires nvcc via CMake. Gate on sys_platform so installs on macOS
    # (MPS) and other non-Linux platforms don't try to build it from source;
    # every fastvideo_kernel import elsewhere is already guarded to degrade
    # gracefully when the package is absent.
    "fastvideo-kernel==0.3.2; sys_platform == 'linux'",
    "wheel",

    # Training Dependencies
    "torchdata",
    "pyarrow",
    "datasets>=4.0.0",
    "av",

    # Preprocessing Dependencies
    "torchcodec",
    "ray>=2.49.1",
    "ftfy>=6.3.1",

    # Job Runner
    "fastapi>=0.129.0",
    "uvicorn>=0.41.0"
]

[tool.uv]
prerelease = "allow"

[tool.uv.workspace]
members = ["apps/dreamverse"]

[tool.uv.sources]
# PyPI's aarch64 fastvideo-kernel wheels do not cover the NVIDIA DGX Spark /
# GB10 (sm_121), so on ARM build from the in-tree source instead; x86_64 keeps
# the published PyPI wheel. This table is uv-only: plain pip ignores it. Pass
# --no-sources to bypass it (e.g. when the kernel is already installed, as in
# the Spark install guide).
fastvideo-kernel = [
  { path = "fastvideo-kernel", marker = "platform_machine == 'aarch64'" },
]
# torch / torchvision / torchaudio are intentionally NOT pinned to a PyTorch
# index here. Pick the supported build explicitly at install time:
#   UV_TORCH_BACKEND=cu126 uv pip install -e ".[dev]"   # CUDA 12
#   UV_TORCH_BACKEND=cu130 uv pip install -e ".[dev]"   # CUDA 13
# uv reads UV_TORCH_BACKEND / --torch-backend (not a pyproject setting). The
# supported backends (cu126, cu130) both provide torch 2.12.0.
imagebind = { git = "https://github.com/facebookresearch/ImageBind.git", rev = "53680b02d7e37b19b124fa37bae4b6c98c38f5be" }
# FA4 cute, pinned to a cutlass-4.5-compatible revision. torch.compile support
# comes from FastVideo's own custom_op wrappers.
flash-attn-4 = { git = "https://github.com/Dao-AILab/flash-attention.git", rev = "82d6441eec5d4dfec120153db2c0145ae855a083", subdirectory = "flash_attn/cute" }

[project.optional-dependencies]

# flash-attn: uv pip install flash-attn==2.8.1 --no-cache-dir --no-build-isolation

swanlab = [
    "swanlab>=0.6.7",
]

lint = [
    "pre-commit==4.0.1",
    "types-aiofiles",
    "types-requests",
]

test = [
    "av",
    "pytorch-msssim>=1.0.0",
    "pytest",
    "pandas",
    "plotly",
]

# Eval extras. See fastvideo/eval/README.md for which install path
# covers which metrics. detectron2 (vbench color/object_class/etc.) is
# CUDA-version-pinned and not cleanly pip-installable — install it
# separately as documented in the README. decord has no aarch64 wheels
# (upstream effectively unmaintained); it's split into its own extra so
# the main eval extras install on ARM systems too. eval/io/video.py has
# a PyAV fallback that covers the video-decode path without decord. The
# judge.* metrics (eval-judge) call a remote API and need a key, so they
# are opt-in and intentionally kept out of the default [eval] set.
eval-vbench     = ["openai-clip", "pyiqa", "easydict"]
eval-physics-iq = []
eval-audio = ["jiwer", "librosa", "pyloudnorm", "hear21passt", "audiobox_aesthetics", "imagebind", "pytorchvideo"]
eval-judge = ["google-genai"]
eval = [
    # ptlflow>=0.4: get_model() takes ckpt_path (0.3 called it
    # pretrained_ckpt); unpinned, the resolver backtracked to 0.2.5.
    "lpips", "ptlflow>=0.4", "qwen-vl-utils",
    "fastvideo[eval-vbench]", "fastvideo[eval-physics-iq]",
]
eval-fast-decode = ["decord"]
eval-full = ["fastvideo[eval]", "fastvideo[eval-audio]", "qwen-omni-utils"]

dev = [ "fastvideo[lint]", "fastvideo[test]", ]

prompt-safety = [
    "fasttext",
]

prompt-enhancer = [
    "httpx",
]

streaming = [
    "fastvideo[prompt-enhancer]",
    "fastvideo[prompt-safety]",
    "websockets",
]

dreamverse = [
    "uvicorn[standard]>=0.41.0",
    "cerebras-cloud-sdk",
    # PyPI forbids direct URL deps in published metadata; pin upstream cute via
    # [tool.uv.sources] above (same pattern as imagebind) so the wheel stays
    # publishable while uv workspace installs still get it.
    "flash-attn-4",
    "flashinfer-python",
    "openai>=1.40",
]

rocm = [
    "amdsmi",
]

[project.scripts]
fastvideo = "fastvideo.entrypoints.cli.main:main"
dreamverse-server = "dreamverse.server_entry:cli"
dreamverse-mock-server = "dreamverse.mock_server:cli"

[tool.setuptools.packages.find]
where = [".", "apps/dreamverse"]
exclude = ["assets*", "docker*", "docs", "scripts*", "apps*", "dreamverse.tests*"]

[tool.setuptools.package-data]
dreamverse = ["prompts/*.md"]

[tool.wheel]
exclude = ["assets*", "docker*", "docs", "scripts*", "apps*"]

[tool.pytest.ini_options]
markers = [
    "gpu: requires real GPU + model weights; skip in CI",
]


[tool.mypy]
warn_unused_configs = true
ignore_missing_imports = true
disallow_untyped_calls = true
check_untyped_defs = true
follow_imports = "silent"

[tool.codespell]
# ``*/_vendored/*`` matches upstream-provenance files vendored under any
# ``_vendored/`` subdir (project-wide convention; mirrors the
# ``_``-prefixed auto-discovery skip).
skip = "./data,./wandb,apps/fastvideo_studio/package-lock.json,apps/performance_dashboard/frontend/package-lock.json,*/_vendored/*"
# "tread" matches daVinci-MagiHuman's acronym "TReAD" (Token Routing and
# Early Drop). codespell lowercases ignore-words entries, so the single
# lowercase form silences all case variants.
ignore-words-list = "tread,passt,dout"

[tool.ruff]
# Allow lines to be as long as 120.
line-length = 120
extend-exclude = [
    # Upstream submodules under fastvideo/third_party/ are not our style
    # territory; modern-dep compat lives as runtime shims in our metric
    # __init__.py files.
    "fastvideo/third_party/**",
]

[tool.ruff.lint]
select = [
    # pycodestyle
    "E",
    # Pyflakes
    "F",
    # pyupgrade
    "UP",
    # flake8-bugbear
    "B",
    # flake8-simplify
    "SIM",
    # isort
    # "I",
    "G",
]
ignore = [
    # star imports
    "F405", "F403",
    # lambda expression assignment
    "E731",
    # Loop control variable not used within loop body
    "B007",
    # f-string format
    "UP032",
    # line too long
    "E501",
]

[tool.ruff.lint.per-file-ignores]
# Ignore all files that end in `_test.py`.
"fastvideo/models/hunyuan/diffusion/pipelines/pipeline_hunyuan_video.py" = ["E741"]


[tool.yapf]
column_limit = 120

[tool.isort]
line_length = 120
use_parentheses = true
skip_gitignore = true


[project.urls]
Repository = "https://github.com/hao-ai-lab/FastVideo"
#  Used by Comfy Registry https://comfyregistry.org

[tool.comfy]
PublisherId = "fastvideo"
DisplayName = "FastVideo"
Icon = "https://raw.githubusercontent.com/hao-ai-lab/FastVideo/main/comfyui/assets/icon_simple.svg"