pyproject.toml

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

# We're not building the project itself
[tool.setuptools]
py-modules = []

[project]
name = "tabbyAPI"
version = "0.0.1"
description = "An OAI compatible exllamav3 API that's both lightweight and fast"
requires-python = ">=3.10"
dependencies = [
    "fastapi-slim >= 0.115",
    "pydantic >= 2.11, < 3",
    "ruamel.yaml",
    "rich",
    "uvicorn >= 0.28.1",
    "jinja2 >= 3.0.0",
    "loguru",
    "sse-starlette >= 2.2.0",
    "packaging",
    "tokenizers >= 0.21.0",
    "numpy",
    "aiofiles",
    "aiohttp",
    "async_lru",
    "huggingface_hub",
    "psutil",
    "httptools >= 0.5.0",
    "pillow",
    "requests",

    # Improved asyncio loops
    "uvloop ; platform_system == 'Linux' and platform_machine == 'x86_64'",
    "winloop ; platform_system == 'Windows'",

    # For python 3.12
    "setuptools ; python_version >= '3.12'"
]

[project.urls]
"Homepage" = "https://github.com/theroyallab/tabbyAPI"

[project.optional-dependencies]
extras = [
    # Heavy dependencies that aren't for everyday use
    "infinity-emb",
    # Match the bound infinity-emb declares for its torch engine
    "sentence-transformers < 4.0",
    # infinity-emb uses the pre-1.0 huggingface_hub API (HfFolder) and doesn't
    # bound it upstream. See issue #414.
    "huggingface_hub < 1.0",
]
dev = [
    "ruff == 0.11.10"
]
cu12 = [
    # Torch (Extra index URLs not support in pyproject.toml)
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp313-cp313-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.13'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp313-cp313-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.13'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp312-cp312-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp311-cp311-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
    "torch @ https://download.pytorch.org/whl/cu128/torch-2.9.0%2Bcu128-cp310-cp310-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",

    # Triton
    "triton ; platform_system == 'Linux' and platform_machine == 'x86_64'",
    "triton-windows ; platform_system == 'Windows'",

    # Flash Linear Attention
    "flash-linear-attention >= 0.5.0 ; (platform_system == 'Linux' and platform_machine == 'x86_64') or platform_system == 'Windows'",

    # Exl3
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp313-cp313-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.13'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp311-cp311-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.11'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp310-cp310-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.10'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.13'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.11'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4+cu128.torch2.9.0-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.10'",
]
cu13 = [
    # Torch (Extra index URLs not support in pyproject.toml)
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp314-cp314-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.14'",
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp313-cp313-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.13'",
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp314-cp314-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.14'",
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp313-cp313-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.13'",
    "torch @ https://download.pytorch.org/whl/cu130/torch-2.11.0%2Bcu130-cp312-cp312-manylinux_2_28_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",

    # Triton
    "triton ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version >= '3.12'",
    "triton-windows ; platform_system == 'Windows' and python_version >= '3.12'",

    # Flash Linear Attention
    "flash-linear-attention >= 0.5.0 ; python_version >= '3.12' and ((platform_system == 'Linux' and platform_machine == 'x86_64') or platform_system == 'Windows')",

    # Exl3
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp314-cp314-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.14'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp313-cp313-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.13'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp312-cp312-win_amd64.whl ; platform_system == 'Windows' and python_version == '3.12'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp314-cp314-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.14'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.13'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Bcu132.torch2.11.0-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
]
rocm = [
    # AMD's pip-native ROCm 10 stack ("TheRock"): the runtime comes from pip, no system
    # ROCm needed, only the kernel driver. These resolve from AMD's indexes
    # (stable.repo.amd.com/rocm/pytorch/whl-next and .../rocm/core/whl-next), which
    # start.py and docker/Dockerfile.rocm pass on the command line (pyproject can't carry
    # an index URL), together with a per-GPU amd-torch-device-gfxNNNN package. The indexes
    # also carry newer lines, so everything is pinned. The torch pinned here is the AMD
    # build the exllamav3 ROCm wheels are built against; bump the two together. Linux
    # only; Windows ROCm stays on a JIT build of exllamav3 for now.
    "torch == 2.13.0+rocm10.0.0 ; platform_system == 'Linux' and platform_machine == 'x86_64'",
    "rocm[libraries] == 10.0.0 ; platform_system == 'Linux' and platform_machine == 'x86_64'",
    # Triton: torch's own pin pulls in the matching ROCm build. flash-linear-attention is
    # not needed: exllamav3 vendors the kernels it uses

    # Exl3. One wheel covers every RDNA generation (gfx10-1, gfx10-3, gfx11 and gfx12
    # generic targets); CDNA (Instinct) needs a JIT build
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Brocm10.0.torch2.13.0-cp314-cp314-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.14'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Brocm10.0.torch2.13.0-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.13'",
    "exllamav3 @ https://github.com/turboderp-org/exllamav3/releases/download/v1.6.4/exllamav3-1.6.4%2Brocm10.0.torch2.13.0-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and platform_machine == 'x86_64' and python_version == '3.12'",
]
# MARK: Ruff options

[tool.ruff]
# Exclude a variety of commonly ignored directories.
exclude = [
    ".git",
    ".git-rewrite",
    ".mypy_cache",
    ".pyenv",
    ".pytest_cache",
    ".ruff_cache",
    ".venv",
    ".vscode",
    "__pypackages__",
    "_build",
    "build",
    "dist",
    "node_modules",
    "site-packages",
    "venv",
    "colab",
    "docker",
]

# For dev/debug purposes
extend-exclude = ["*_test_.py"]

# Same as Black.
line-length = 100
indent-width = 4

# Assume Python 3.10
target-version = "py310"

[tool.ruff.lint]
# Enable preview
preview = true

# Enable Pyflakes (`F`) and a subset of the pycodestyle (`E`)  codes by default.
# Unlike Flake8, Ruff doesn't enable pycodestyle warnings (`W`) or
# McCabe complexity (`C901`) by default.
# Enable flake8-bugbear (`B`) rules, in addition to the defaults.
select = ["E4", "E7", "E9", "F", "B"]
extend-select = [
    "D419",  # empty-docstring
    "PLC2401",  # non-ascii-name
    "E501",  # line-too-long
    "W291",  # trailing-whitespace
    "PLC0414",  # useless-import-alias
    "PLE0101",  # return-in-init
    "F706",  # return-outside-function
    "F704",  # yield-outside-function
    "PLE0116",  # continue-in-finally
    "PLE0117",  # nonlocal-without-binding
    "PLE0241",  # duplicate-bases
    "PLE0302",  # unexpected-special-method-signature
    "PLE0604",  # invalid-all-object
    "PLE0704",  # misplaced-bare-raise
    "PLE1205",  # logging-too-many-args
    "PLE1206",  # logging-too-few-args
    "PLE1307",  # bad-string-format-type
    "PLE1310",  # bad-str-strip-call
    "PLE1507",  # invalid-envvar-value
    "PLR0124",  # comparison-with-itself
    "PLR0202",  # no-classmethod-decorator
    "PLR0203",  # no-staticmethod-decorator
    "PLR0206",  # property-with-parameters
    "PLR1704",  # redefined-argument-from-local
    "PLR1711",  # useless-return
    "C416",  # unnecessary-comprehension
    "PLW0108",  # unnecessary-lambda
    "PLW0127",  # self-assigning-variable
    "PLW0129",  # assert-on-string-literal
    "PLW0602",  # global-variable-not-assigned
    "PLW0604",  # global-at-module-level
    "F401",  # unused-import
    "F841",  # unused-variable
    "E722",  # bare-except
    "PLW0711",  # binary-op-exception
    "PLW1501",  # bad-open-mode
    "PLW1508",  # invalid-envvar-default
    "PLW1509",  # subprocess-popen-preexec-fn
]
ignore = [
    "PLR6301",  # no-self-use
    "UP004",  # useless-object-inheritance
    "PLR0904",  # too-many-public-methods
    "PLR0911",  # too-many-return-statements
    "PLR0912",  # too-many-branches
    "PLR0913",  # too-many-arguments
    "PLR0914",  # too-many-locals
    "PLR0915",  # too-many-statements
    "PLR0916",  # too-many-boolean-expressions
    "PLW0120",  # useless-else-on-loop
    "PLW0406",  # import-self
    "PLW0603",  # global-statement
    "PLW1641",  # eq-without-hash
]

# Allow fix for all enabled rules (when `--fix`) is provided.
fixable = ["ALL"]
unfixable = ["B"]

# Allow unused variables when underscore-prefixed.
dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"

[tool.ruff.format]
# Like Black, use double quotes for strings.
quote-style = "double"

# Like Black, indent with spaces, rather than tabs.
indent-style = "space"

# Like Black, respect magic trailing commas.
skip-magic-trailing-comma = false

# Like Black, automatically detect the appropriate line ending.
line-ending = "auto"

# Don't format/validate test scripts
[tool.ruff.lint.per-file-ignores]
"tests/**" = ["ALL"]