pyproject.toml

[build-system]
requires = ["hatchling"]
build-backend = "hatchling.build"

[project]
name = "distilabel"
description = "Distilabel is an AI Feedback (AIF) framework for building datasets with and for LLMs."
readme = "README.md"
requires-python = ">=3.9"
license = "Apache-2.0"
keywords = ["llm", "annotation", "alignment", "synthetic", "data", "rlaif"]
authors = [{ name = "Argilla", email = "admin@argilla.io" }]
classifiers = [
    "Development Status :: 4 - Beta",
    "Programming Language :: Python",
    "Programming Language :: Python :: 3.9",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: Implementation :: CPython",
    "Programming Language :: Python :: Implementation :: PyPy",
]
dependencies = [
    # Bump `datasets` to support `load_dataset` from cache
    # Ref https://github.com/huggingface/datasets/releases/tag/2.16.0
    "datasets >= 2.16.0",
    "httpx >= 0.25.2",
    "importlib-resources >= 6.1.1; python_version < '3.9'",
    "Jinja2 >= 3.1.2",
    "multiprocess >= 0.70",
    "nest-asyncio >= 1.6.0",
    "networkx >= 3.0",
    "pydantic >= 2.0",
    "rich >= 13.5.0",
    "scipy >= 1.10.0",
    "typer >= 0.9.0",
    "tblib >= 3.0.0",
    "orjson >= 3.10.0",
    "universal_pathlib >= 0.2.2",
    "portalocker >= 2.8.2",
]
dynamic = ["version"]

[project.scripts]
distilabel = "distilabel.cli.app:app"

[project.entry-points."mkdocs.plugins"]
"distilabel/components-gallery" = "distilabel.utils.mkdocs.components_gallery:ComponentsGalleryPlugin"

[project.optional-dependencies]
dev = ["ruff == 0.6.2", "pre-commit >= 3.5.0"]
docs = [
    "mkdocs-material >=9.5.17",
    "mkdocstrings[python] >= 0.24.0",
    "mkdocs-literate-nav >= 0.6.1",
    "mkdocs-section-index >= 0.3.8",
    "mkdocs-gen-files >= 0.5.0",
    "mkdocs-glightbox >= 0.4.0",
    "material-plausible-plugin>=0.2.0",
    "mike >= 2.0.0",
    "Pillow >= 9.5.0",
    "CairoSVG >= 2.7.1",
    "mknotebooks >= 0.8.0",
    "pandas >= 2.0",
    "tabulate>=0.9.0",
]
tests = [
    "pytest >= 7.4.0",
    "pytest-asyncio",
    "nest-asyncio",
    "pytest-timeout",
    "pytest-codspeed",
]

# Optional LLMs, integrations, etc
anthropic = ["anthropic >= 0.20.0"]
argilla = ["argilla >= 2.0.0", "ipython"]
cohere = ["cohere >= 5.2.0"]
groq = ["groq >= 0.4.1"]
hf-inference-endpoints = ["huggingface_hub >= 0.22.0"]
hf-transformers = ["transformers >= 4.34.1", "torch >= 2.0.0"]
instructor = ["instructor >= 1.2.3"]
litellm = ["litellm >= 1.30.0"]
llama-cpp = ["llama-cpp-python >= 0.2.0"]
mistralai = ["mistralai >= 1.0.0"]
ollama = ["ollama >= 0.1.7"]
openai = ["openai >= 1.0.0"]
outlines = ["outlines >= 0.0.40", "numba >= 0.54.0"]
ray = ["ray[default] >= 2.31.0"]
vertexai = ["google-cloud-aiplatform >= 1.38.0"]
vllm = [
    "vllm >= 0.5.3",
    "filelock >= 3.13.4",
    # `setuptools` is needed to be installed if installed with `uv pip install distilabel[vllm]`
    "setuptools",
]
sentence-transformers = ["sentence-transformers >= 3.0.0"]
faiss-cpu = ["faiss-cpu >= 1.8.0"]
faiss-gpu = ["faiss-gpu >= 1.7.2"]
text-clustering = [
    "umap-learn >= 0.5.6",
    "scikit-learn >= 1.4.1",
    "matplotlib >= 3.8.3",   # For the figure (even though it's optional)
]

# minhash
minhash = ["datasketch >= 1.6.5", "nltk>3.8.1"]

[project.urls]
Documentation = "https://distilabel.argilla.io/"
Issues = "https://github.com/argilla/distilabel/issues"
Source = "https://github.com/argilla/distilabel"

[tool.hatch.version]
path = "src/distilabel/__init__.py"

[tool.ruff]
line-length = 88
exclude = ["docs"]

[tool.ruff.lint]
select = ["E", "W", "F", "I", "C", "B"]
ignore = ["E501", "B905", "B008"]

[tool.pytest.ini_options]
testpaths = ["tests"]