pyproject.toml

[build-system]
requires = ["setuptools>=77", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "diffsynth"
dynamic = ["version"]
description = "Enjoy the magic of Diffusion models!"
readme = "README.md"
authors = [{name = "ModelScope Team"}]
license = "Apache-2.0"
license-files = ["LICENSE"]
requires-python = ">=3.10.1"
dependencies = [
    "torch==2.6.0",
    "torchvision",
    "transformers",
    "imageio[ffmpeg]",
    "safetensors",
    "einops",
    "modelscope",
    "ftfy",
    "pandas",
    "accelerate",
    "peft",
]
classifiers = [
    "Programming Language :: Python :: 3",
    "Operating System :: OS Independent",
]

[tool.setuptools.packages.find]
where = ["./"]
include = ["diffsynth", "diffsynth.*"]

[tool.setuptools.dynamic]
version = {attr = "diffsynth.version.__version__"}

[project.optional-dependencies]
audio = [
    "av",
    "torchaudio",
    "torchcodec",
    "librosa"
]
quant = [
    "bitsandbytes",
    "comfy-kitchen",
    "torchao>=0.16"
]
training = [
    "deepspeed<0.19"
]
logger = [
    "tensorboard",
    "swanlab",
    "wandb"
]
npu = [
    "torch==2.7.1+cpu",
    "torch-npu==2.7.1",
    "torchvision==0.22.1+cpu"
]
npu_aarch64 = [
    "torch==2.7.1",
    "torch-npu==2.7.1",
    "torchvision==0.22.1"
]
infiniteyou = [
    "insightface",
    "facexlib"
]
ses = [
    "pywt"
]
nexusgen = [
    "qwen_vl_utils",
    "transformers==4.49.0"
]
all = [
    "av",
    "torchaudio",
    "torchcodec",
    "librosa",
    "bitsandbytes",
    "comfy-kitchen",
    "torchao>=0.16",
    "deepspeed<0.19",
    "tensorboard",
    "swanlab",
    "wandb"
]

[tool.setuptools]
include-package-data = true