updates
This commit is contained in:
@@ -7,7 +7,7 @@ name = "axolotl"
|
|||||||
dynamic = ["version"]
|
dynamic = ["version"]
|
||||||
description = "LLM Trainer"
|
description = "LLM Trainer"
|
||||||
readme = "README.md"
|
readme = "README.md"
|
||||||
requires-python = ">=3.10"
|
requires-python = ">=3.10,<3.13"
|
||||||
license = {text = "Apache-2.0"}
|
license = {text = "Apache-2.0"}
|
||||||
authors = [
|
authors = [
|
||||||
{name = "Axolotl AI"},
|
{name = "Axolotl AI"},
|
||||||
@@ -25,7 +25,6 @@ classifiers = [
|
|||||||
]
|
]
|
||||||
|
|
||||||
dependencies = [
|
dependencies = [
|
||||||
# Core dependencies
|
|
||||||
"torch>=2.6.0",
|
"torch>=2.6.0",
|
||||||
"packaging>=23.2",
|
"packaging>=23.2",
|
||||||
"huggingface_hub>=0.33.0",
|
"huggingface_hub>=0.33.0",
|
||||||
@@ -38,16 +37,12 @@ dependencies = [
|
|||||||
"hf_xet==1.1.5",
|
"hf_xet==1.1.5",
|
||||||
"kernels==0.9.0",
|
"kernels==0.9.0",
|
||||||
"trackio",
|
"trackio",
|
||||||
|
|
||||||
# Optimization and training
|
|
||||||
"optimum==1.16.2",
|
"optimum==1.16.2",
|
||||||
"hf_transfer",
|
"hf_transfer",
|
||||||
"sentencepiece",
|
"sentencepiece",
|
||||||
"gradio==5.41.1",
|
"gradio==5.41.1",
|
||||||
|
|
||||||
# Infrastructure
|
|
||||||
"modal==1.0.2",
|
"modal==1.0.2",
|
||||||
"pydantic==2.10.6",
|
"pydantic>=2.10.6",
|
||||||
"addict",
|
"addict",
|
||||||
"fire",
|
"fire",
|
||||||
"PyYAML>=6.0",
|
"PyYAML>=6.0",
|
||||||
@@ -57,8 +52,6 @@ dependencies = [
|
|||||||
"colorama",
|
"colorama",
|
||||||
"numba",
|
"numba",
|
||||||
"numpy>=1.24.4,<3.0",
|
"numpy>=1.24.4,<3.0",
|
||||||
|
|
||||||
# QLora dependencies
|
|
||||||
"evaluate==0.4.1",
|
"evaluate==0.4.1",
|
||||||
"scipy",
|
"scipy",
|
||||||
"scikit-learn>=1.7.0",
|
"scikit-learn>=1.7.0",
|
||||||
@@ -66,33 +59,23 @@ dependencies = [
|
|||||||
"art",
|
"art",
|
||||||
"tensorboard",
|
"tensorboard",
|
||||||
"python-dotenv==1.0.1",
|
"python-dotenv==1.0.1",
|
||||||
|
|
||||||
# Remote filesystems
|
|
||||||
"s3fs>=2024.5.0",
|
"s3fs>=2024.5.0",
|
||||||
"gcsfs>=2024.5.0",
|
"gcsfs>=2024.5.0",
|
||||||
"adlfs>=2024.5.0",
|
"adlfs>=2024.5.0",
|
||||||
"ocifs==1.3.2",
|
"ocifs==1.3.2",
|
||||||
|
|
||||||
# Other utilities
|
|
||||||
"zstandard>=0.23.0",
|
"zstandard>=0.23.0",
|
||||||
"fastcore",
|
"fastcore",
|
||||||
|
|
||||||
# LM eval harness
|
|
||||||
"lm_eval==0.4.7",
|
"lm_eval==0.4.7",
|
||||||
"langdetect==1.0.9",
|
"langdetect==1.0.9",
|
||||||
"immutabledict==4.2.0",
|
"immutabledict==4.2.0",
|
||||||
"antlr4-python3-runtime==4.13.2",
|
"antlr4-python3-runtime==4.13.2",
|
||||||
|
|
||||||
# Training optimizers
|
|
||||||
"schedulefree==1.4.1",
|
"schedulefree==1.4.1",
|
||||||
|
"mistral-common==1.8.3",
|
||||||
|
|
||||||
# Axolotl contribs
|
# Axolotl contribs
|
||||||
"axolotl-contribs-lgpl @ git+https://github.com/axolotl-ai-cloud/axolotl-contribs-lgpl.git@numpy",
|
"axolotl-contribs-lgpl @ git+https://github.com/axolotl-ai-cloud/axolotl-contribs-lgpl.git@numpy",
|
||||||
"axolotl-contribs-mit==0.0.5",
|
"axolotl-contribs-mit==0.0.5",
|
||||||
|
|
||||||
# Mistral
|
|
||||||
"mistral-common==1.8.3",
|
|
||||||
|
|
||||||
# Platform-specific dependencies (Linux by default, excluded on macOS)
|
# Platform-specific dependencies (Linux by default, excluded on macOS)
|
||||||
"triton>=3.0.0 ; sys_platform != 'darwin'",
|
"triton>=3.0.0 ; sys_platform != 'darwin'",
|
||||||
"xformers>=0.0.23.post1 ; sys_platform != 'darwin'",
|
"xformers>=0.0.23.post1 ; sys_platform != 'darwin'",
|
||||||
@@ -112,31 +95,11 @@ ring-flash-attn = [
|
|||||||
"ring-flash-attn>=0.1.7",
|
"ring-flash-attn>=0.1.7",
|
||||||
"yunchang==0.6.0",
|
"yunchang==0.6.0",
|
||||||
]
|
]
|
||||||
|
mamba-ssm = ["mamba-ssm>=2.2.0", "causal_conv1d>=1.4.0",]
|
||||||
mamba-ssm = [
|
auto-gptq = ["auto-gptq==0.5.1"]
|
||||||
"mamba-ssm>=2.2.0",
|
mlflow = ["mlflow"]
|
||||||
"causal_conv1d>=1.4.0",
|
galore = ["galore_torch"]
|
||||||
]
|
apollo = ["apollo-torch"]
|
||||||
|
|
||||||
# Quantization
|
|
||||||
auto-gptq = [
|
|
||||||
"auto-gptq==0.5.1",
|
|
||||||
]
|
|
||||||
|
|
||||||
# Experiment tracking
|
|
||||||
mlflow = [
|
|
||||||
"mlflow",
|
|
||||||
]
|
|
||||||
|
|
||||||
# Optimizers
|
|
||||||
galore = [
|
|
||||||
"galore_torch",
|
|
||||||
]
|
|
||||||
|
|
||||||
apollo = [
|
|
||||||
"apollo-torch",
|
|
||||||
]
|
|
||||||
|
|
||||||
optimizers = [
|
optimizers = [
|
||||||
"galore_torch",
|
"galore_torch",
|
||||||
"apollo-torch",
|
"apollo-torch",
|
||||||
@@ -144,24 +107,10 @@ optimizers = [
|
|||||||
"torch-optimi==0.2.1",
|
"torch-optimi==0.2.1",
|
||||||
"came_pytorch==0.1.3",
|
"came_pytorch==0.1.3",
|
||||||
]
|
]
|
||||||
|
ray = ["ray[train]"]
|
||||||
# Distributed training
|
vllm = ["vllm>=0.10.0"]
|
||||||
ray = [
|
llmcompressor = ["llmcompressor>=0.5.1"]
|
||||||
"ray[train]",
|
|
||||||
]
|
|
||||||
|
|
||||||
# Inference (Note: vllm and llmcompressor have conflicting dependencies, install separately)
|
|
||||||
vllm = [
|
|
||||||
"vllm>=0.10.0",
|
|
||||||
]
|
|
||||||
|
|
||||||
llmcompressor = [
|
|
||||||
"llmcompressor>=0.5.1",
|
|
||||||
]
|
|
||||||
|
|
||||||
# Development dependencies
|
|
||||||
dev = [
|
dev = [
|
||||||
# Testing
|
|
||||||
"pytest",
|
"pytest",
|
||||||
"pytest-cov",
|
"pytest-cov",
|
||||||
"pytest-retry",
|
"pytest-retry",
|
||||||
@@ -170,19 +119,13 @@ dev = [
|
|||||||
"codecov",
|
"codecov",
|
||||||
"codecov-cli",
|
"codecov-cli",
|
||||||
"tbparse",
|
"tbparse",
|
||||||
|
|
||||||
# Code quality
|
|
||||||
"black",
|
"black",
|
||||||
"ruff",
|
"ruff",
|
||||||
"mypy",
|
"mypy",
|
||||||
"pre-commit",
|
"pre-commit",
|
||||||
"types-requests",
|
"types-requests",
|
||||||
|
|
||||||
# Documentation
|
|
||||||
"quartodoc",
|
"quartodoc",
|
||||||
"jupyter",
|
"jupyter",
|
||||||
|
|
||||||
# Utilities
|
|
||||||
"blobfile",
|
"blobfile",
|
||||||
"tiktoken",
|
"tiktoken",
|
||||||
]
|
]
|
||||||
@@ -214,7 +157,7 @@ line-length = 88
|
|||||||
target-version = "py310"
|
target-version = "py310"
|
||||||
|
|
||||||
[tool.ruff.lint]
|
[tool.ruff.lint]
|
||||||
select = ["E", "F", "W", "C90", "B"]
|
select = ["E", "F", "W", "C90", "B", "I"]
|
||||||
ignore = [
|
ignore = [
|
||||||
"E203", # Whitespace before ':'
|
"E203", # Whitespace before ':'
|
||||||
"E501", # Line too long
|
"E501", # Line too long
|
||||||
@@ -279,6 +222,9 @@ flash-attn = [
|
|||||||
"setuptools",
|
"setuptools",
|
||||||
{ requirement = "torch", match-runtime = true },
|
{ requirement = "torch", match-runtime = true },
|
||||||
]
|
]
|
||||||
|
auto-gptq = [
|
||||||
|
{ requirement = "torch", match-runtime = true },
|
||||||
|
]
|
||||||
autoawq = ["torch"]
|
autoawq = ["torch"]
|
||||||
triton = ["torch"]
|
triton = ["torch"]
|
||||||
bitsandbytes = ["torch"]
|
bitsandbytes = ["torch"]
|
||||||
|
|||||||
@@ -27,7 +27,7 @@ version_tuple: VERSION_TUPLE
|
|||||||
commit_id: COMMIT_ID
|
commit_id: COMMIT_ID
|
||||||
__commit_id__: COMMIT_ID
|
__commit_id__: COMMIT_ID
|
||||||
|
|
||||||
__version__ = version = "0.12.1.dev87+g031a2474c.d20250911"
|
__version__ = version = "0.12.1.dev88+gc42ba73d4.d20250915"
|
||||||
__version_tuple__ = version_tuple = (0, 12, 1, "dev87", "g031a2474c.d20250911")
|
__version_tuple__ = version_tuple = (0, 12, 1, "dev88", "gc42ba73d4.d20250915")
|
||||||
|
|
||||||
__commit_id__ = commit_id = "g031a2474c"
|
__commit_id__ = commit_id = "gc42ba73d4"
|
||||||
|
|||||||
Reference in New Issue
Block a user