[project]
name = "next-token-prediction"
version = "1.0.0"
description = "Interactive next-token prediction workbench"
readme = "README.md"
requires-python = ">=3.11,<3.12"
dependencies = [
    "fastapi==0.137.2",
    "huggingface-hub==0.36.2",
    "torch==2.3.1+cpu",
    "transformers==4.44.2",
    "uvicorn==0.49.0",
]

[project.optional-dependencies]
gguf = [
    "llama-cpp-python==0.3.30",
]

[dependency-groups]
dev = [
    "httpx>=0.27.0",
    "pytest>=8.0.0",
]

[tool.uv]
package = false

[tool.uv.sources]
torch = { index = "pytorch-cpu" }
llama-cpp-python = { index = "llama-cpp-cpu" }

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

[[tool.uv.index]]
name = "llama-cpp-cpu"
url = "https://abetlen.github.io/llama-cpp-python/whl/cpu"
explicit = true

[tool.pytest.ini_options]
pythonpath = ["."]