pyproject.toml

[project]
name = "modernvbert"
version = "1.0.0"
description = "ModernVBERT is a 250M-parameter vision–language encoder that aligns a text-encoder (Ettin-150M) with a vision-encoder (SigLIP2-B) through a MLM objective. When fine-tuned for document retrieval, ModernVBERT sets a new state of the art for sub-1B models on ViDoRe tasks."
readme = "README.md"
requires-python = ">=3.10"
dependencies = [
    "torch>=2.2.0",
    "pillow>=11.3.0",
    "transformers>=4.51.0",
    "numpy>=1.24.0,<2.0.0",
    "google-genai>=1.43.0",
]

[tool.setuptools]
packages = ["modernvbert"]
package-dir = {"" = "src"}

[tool.pytest.ini_options]
filterwarnings = ["ignore::Warning"]
testpaths = ["tests"]

[project.optional-dependencies]
dev = [
    "pytest>=7.0",
]