QuantLLM/pyproject.toml at main · codewithdark-git/QuantLLM · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
[build-system]
requires = ["setuptools>=68", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "quantllm"
version = "2.1.0"
description = "Ultra-fast LLM quantization, fine-tuning, and deployment with one line of code"
readme = "README.md"
license = {text = "MIT"}
authors = [
    {name = "Dark Coder", email = "codewithdark90@gmail.com"},
]
requires-python = ">=3.10"
classifiers = [
    "Development Status :: 4 - Beta",
    "Intended Audience :: Science/Research",
    "Intended Audience :: Developers",
    "License :: OSI Approved :: MIT License",
    "Operating System :: OS Independent",
    "Programming Language :: Python :: 3",
    "Programming Language :: Python :: 3.10",
    "Programming Language :: Python :: 3.11",
    "Programming Language :: Python :: 3.12",
    "Programming Language :: Python :: 3.13",
    "Topic :: Scientific/Engineering :: Artificial Intelligence",
    "Topic :: Software Development :: Libraries :: Python Modules",
]
keywords = [
    "llm", "quantization", "gguf", "transformers", "pytorch",
    "fine-tuning", "lora", "inference", "llama", "mistral",
    "machine-learning", "deep-learning", "nlp",
]
dependencies = [
    "torch>=2.0.0",
    "transformers>=4.36.0",
    "datasets>=2.14.0",
    "accelerate>=0.24.0",
    "peft>=0.6.0",
    "bitsandbytes>=0.41.0",
    "scipy>=1.10.0",
    "scikit-learn>=1.3.0",
    "tqdm>=4.65.0",
    "numpy>=1.24.0",
    "sentencepiece>=0.1.99",
    "protobuf>=3.20.0",
    "einops>=0.7.0",
    "psutil>=5.9.0",
    "huggingface-hub>=0.19.0",
    "rich>=13.0.0",
    "gguf>=0.10.0",
]

[project.optional-dependencies]
gguf = ["gguf>=0.10.0"]
onnx = [
    "onnx>=1.14.0",
    "onnxruntime>=1.16.0",
    "optimum[onnxruntime]>=1.14.0",
    "onnxscript>=0.1.0",
]
mlx = ["mlx>=0.1.0", "mlx-lm>=0.1.0"]
triton = ["triton>=2.1.0"]
flash = ["flash-attn>=2.3.0"]
hub = ["hf-lifecycle @ git+https://github.com/codewithdark-git/huggingface-lifecycle.git"]
full = [
    "gguf>=0.10.0",
    "onnx>=1.14.0",
    "onnxruntime>=1.16.0",
    "optimum[onnxruntime]>=1.14.0",
    "onnxscript>=0.1.0",
    "triton>=2.1.0",
    "wandb>=0.15.0",
    "tensorboard>=2.14.0",
    "evaluate>=0.4.0",
    "flash-attn>=2.3.0",
]
dev = [
    "pytest>=7.4.0",
    "pytest-cov>=4.1.0",
    "black>=23.0.0",
    "isort>=5.12.0",
    "flake8>=6.1.0",
    "mypy>=1.8.0",
    "sphinx>=7.0.0",
    "sphinx-rtd-theme>=1.3.0",
    "sphinx-copybutton>=0.5.2",
    "myst-parser>=2.0.0",
]

[project.urls]
Homepage = "https://github.com/codewithdark-git/QuantLLM"
Documentation = "https://quantllm.readthedocs.io/"
"Bug Tracker" = "https://github.com/codewithdark-git/QuantLLM/issues"
Sponsor = "https://github.com/sponsors/codewithdark-git"

[project.scripts]
quantllm = "quantllm.cli:main"

[tool.setuptools.packages.find]
exclude = ["tests", "tests.*", "docs", "examples"]

[tool.ruff]
line-length = 120
target-version = "py310"
extend-exclude = [
    "build",
    "dist",
    ".venv",
    "examples/output",
    "docs/_build",
]

[tool.ruff.lint]
# Minimal "blocker" ruleset enforced on every push. The codebase predates
# ruff, so we deliberately start small -- this catches genuine bugs (syntax
# errors, undefined names, broken comparisons) without producing a giant
# reformatting diff. Stricter rules can be opted-in incrementally.
select = [
    "E9",     # pycodestyle runtime errors (syntax, indentation)
    "F63",    # invalid `is` comparisons / `not in` issues
    "F7",     # syntax errors in expressions
    "F82",    # undefined names actually used
    "F811",   # redefinition of unused name
    "F821",   # undefined-name reference
    "F823",   # local variable referenced before assignment
    # NOTE: F841 (unused local) is intentionally NOT enabled yet. The existing
    # codebase has several intentional unused-binding patterns (progress
    # tasks, documenting intent, etc.) and fixing them is out of scope.
    "B006",   # mutable default arguments (genuine bug class)
    "B017",   # ``assertRaises(Exception)`` -- masks real failures
]
ignore = [
    "E501",   # line length: not enforced strictly
    "E402",   # module-level import not at top of file
    "E741",   # ambiguous variable names
]

[tool.myft.lint.per-file-ignores]
"tests/*" = ["F401", "F811", "F841"]
"quantllm/__init__.py" = ["F401", "E402"]
"quantllm/core/__init__.py" = ["F401"]

[tool.mypy]
python_version = "3.10"
ignore_missing_imports = true
disallow_untyped_defs = false
warn_unused_ignores = false

[tool.pytest.ini_options]
minversion = "7.0"
testpaths = ["tests"]
addopts = "-ra --strict-markers"
filterwarnings = [
    "ignore::DeprecationWarning",
    "ignore::FutureWarning",
]