pyproject.toml

[project]
name = "chc-folk-vocabulary"
version = "0.1.0"
description = "Corpus-grounded folk vocabulary for the narrow abilities of the Cattell-Horn-Carroll (CHC) model: retrieval, relevance filtering, descriptor extraction and per-ability fingerprints"
readme = "README.md"
license = "MIT"
authors = [{ name = "Jiho Noh" }, { name = "Shwetaben Ramani" }]
requires-python = ">=3.10"
dependencies = [
    "torch>=2.0",
    "numpy",
    "pyyaml",
    "tqdm",
    "requests>=2.34.2",
    "zstandard>=0.22",
    "nltk>=3.9",
    "faiss-cpu>=1.8",
    "FlagEmbedding>=1.2",
    "accelerate>=0.26",
    # transformers >=4.54 references torch.float8_e8m0fnu, which needs torch>=2.7;
    # this project is pinned to torch 2.6 (cu124), so cap transformers below that.
    "transformers>=4.45,<4.50",
    # DeBERTa-v3 tokenizer (relevance classifier backbone) is SentencePiece-based.
    "sentencepiece>=0.2",
    "protobuf",
    # LLM annotation step (scripts/p3_annotate.py); OpenAI-compatible, honors OPENAI_BASE_URL.
    "openai>=1.0",
]

[project.optional-dependencies]
dev = [
    "pytest",
    "pytest-cov",
    "jupyter",
    "matplotlib",
    "ipykernel",
]
# Statistics, clustering and figures (p3 stats/detectability, p5, p6 scripts).
viz = [
    "scipy",
    "scikit-learn",
    "matplotlib",
    "umap-learn",
    "adjustText",
]

[project.urls]
Repository = "https://github.com/Jiho-YesNLP/chc-folk-vocabulary"
Explorer = "https://chcfolkvocab.yesnlp.us"

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

[tool.hatch.build.targets.wheel]
packages = ["src"]

[[tool.uv.index]]
name = "pytorch-cu124"
url = "https://download.pytorch.org/whl/cu124"
explicit = true

[tool.uv.sources]
# CUDA build on Linux; on macOS/other, resolve CPU torch from PyPI
# so `uv` can build a local env for the viz step (scripts/p5_cluster_viz.py — no GPU).
torch = [
    { index = "pytorch-cu124", marker = "sys_platform == 'linux'" },
]