pyproject.toml

[tool.poetry]
name = "filecatcher"
version = "0.1.0"
description = "A simple package to process and load files into vector databases."
authors = ["htagourti <htagourti@linagora.com>"]
readme = "README.md"
packages = [{ include = "filecatcher" }]

[tool.poetry.dependencies]
python = ">=3.12,<3.13"


loguru = "^0.7.2"
pyyaml = "^6.0.2"
langchain-core = "^0.3.17"
langchain-community = "^0.3.7"
langchain-openai = "^0.2.8"
ragatouille = "^0.0.8.post4"
whisperx = {git = "https://github.com/federicotorrielli/BetterWhisperX", branch="main"}
pyannote-audio = "3.1.1"
ctranslate2 = "4.4.0"
qdrant-client = "^1.12.1"
langchain-qdrant = "^0.2.0"
pydub = "^0.25.1"
pymupdf4llm = "^0.0.17"
aiopath = "^0.7.7"
langchain-experimental = "^0.3.3"
fastembed-gpu = "^0.4.2"
hydra-core = "^1.3.2"
python-dotenv = "^1.0.1"
unstructured = "^0.16.5"
python-docx = "^1.1.2"
pypandoc = "^1.14"
click = "^8.1.7"
git-filter-repo = "^2.45.0"
img2table = "^1.4.0"


[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"