pyproject.toml

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

[project]
name = "eleuther-ccs"
description = "Unsupervised methods of eliciting latent knowledge related to Contrast Consistent Search (CCS)"
readme = "README.md"
requires-python = ">=3.10"
keywords = ["nlp", "interpretability", "language-models", "explainable-ai"]
license = { text = "MIT License" }
dependencies = [
    # Allows us to use device_map in from_pretrained. Also needed for 8bit
    "accelerate",
    # Already a dependency of datasets, but newer versions introduce breaking changes
    "pyarrow==12.0.0",
    # For pseudolabel and prompt normalization. We're picky about the version because
    # the package isn't guaranteed to be stable yet.
    "concept-erasure==0.1.0",
    # Added distributed.split_dataset_by_node for IterableDatasets
    "datasets>=2.9.0",
    "einops",
    # Introduced numpy.typing module
    "numpy>=1.20.0",
    # For logging. Indirectly required by datasets, but just to be safe we specify it here.
    "pandas",
    # Basically any version should work as long as it supports the user's CUDA version
    "pynvml",
    # We upstreamed bugfixes for Literal types in 0.1.1
    "simple-parsing>=0.1.1",
    # Version 1.11 introduced Fully Sharded Data Parallel, which we plan to use soon
    "torch>=1.11.0",
    # Doesn't really matter but versions < 4.0 are very very old (pre-2016)
    "tqdm>=4.0.0",
    # 4.0 introduced the breaking change of using return_dict=True by default
    "transformers>=4.0.0",
    # required by promptsource, which doesn't specify a version
    "jinja2",
    # For visualization of results
    "plotly==5.14.1",
    "kaleido==0.2.1",
    "rich==13.3.5"

]
version = "0.1.1"

[project.optional-dependencies]
dev = [
    "hypothesis",
    "pre-commit",
    "pytest",
    "pyright==1.1.304",
    "scikit-learn",
    "pyfakefs"
]
8bit = [
    "bitsandbytes",
]

[project.scripts]
ccs = "ccs.__main__:run"

[tool.pyright]
include = ["ccs*"]
reportPrivateImportUsage = false

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

[tool.setuptools.package-data]
ccs = ["*.json", "*.yaml"]

[tool.setuptools.packages.find]
include = ["ccs"]

[tool.ruff]
# Enable pycodestyle (`E`), Pyflakes (`F`), and isort (`I`) codes
# See https://beta.ruff.rs/docs/rules/ for more possible rules
select = ["E", "F", "I"]
# Same as Black.
line-length = 88
# Avoid automatically removing unused imports in __init__.py files.
# Such imports will be flagged with a dedicated message suggesting
# that the import is either added to the module's __all__ symbol
ignore-init-module-imports = true