Compare commits
28 commits
main
...
bump-versi
| Author | SHA1 | Date | |
|---|---|---|---|
|
f5b3a673ee | ||
|
|
27ec6e5af2 | ||
|
|
67678d2c29 | ||
|
|
e568000a92 | ||
|
|
6700dd60f0 | ||
|
|
abf578327c | ||
|
|
a4ae80cd6a | ||
|
|
ec47b2984d | ||
|
|
d9d1a63397 | ||
|
|
0bb6379aad | ||
|
|
1c608e8ff7 | ||
|
|
f9adf6834d | ||
|
|
1fd5853741 | ||
|
|
93a70fbe4e | ||
|
|
8b77451e75 | ||
|
|
481f0618ff | ||
|
|
ddf6f6d1f9 | ||
|
|
dddc9eac92 | ||
|
|
da81c94510 | ||
|
|
4c139503c1 | ||
|
|
9d1e3c38bc | ||
|
|
8bc26f4e1d | ||
|
|
044e67f5aa | ||
|
|
ba36c12240 | ||
|
|
0a5652281f | ||
|
|
bf63f79414 | ||
|
|
2973bea3d0 | ||
|
|
ddbbfe52cf |
107 changed files with 1633 additions and 1866 deletions
10
install.ps1
10
install.ps1
|
|
@ -681,13 +681,13 @@ shell.Run cmd, 0, False
|
||||||
if ($SkipTorch) {
|
if ($SkipTorch) {
|
||||||
# No-torch: install unsloth + unsloth-zoo with --no-deps, then
|
# No-torch: install unsloth + unsloth-zoo with --no-deps, then
|
||||||
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
||||||
uv pip install --python $VenvPython --no-deps --reinstall-package unsloth --reinstall-package unsloth-zoo "unsloth>=2026.3.14" unsloth-zoo
|
uv pip install --python $VenvPython --no-deps --reinstall-package unsloth --reinstall-package unsloth-zoo "unsloth>=2026.3.16" unsloth-zoo
|
||||||
$NoTorchReq = Find-NoTorchRuntimeFile
|
$NoTorchReq = Find-NoTorchRuntimeFile
|
||||||
if ($NoTorchReq) {
|
if ($NoTorchReq) {
|
||||||
uv pip install --python $VenvPython --no-deps -r $NoTorchReq
|
uv pip install --python $VenvPython --no-deps -r $NoTorchReq
|
||||||
}
|
}
|
||||||
} else {
|
} else {
|
||||||
uv pip install --python $VenvPython --reinstall-package unsloth --reinstall-package unsloth-zoo "unsloth>=2026.3.14" unsloth-zoo
|
uv pip install --python $VenvPython --reinstall-package unsloth --reinstall-package unsloth-zoo "unsloth>=2026.3.16" unsloth-zoo
|
||||||
}
|
}
|
||||||
if ($StudioLocalInstall) {
|
if ($StudioLocalInstall) {
|
||||||
Write-Host "==> Overlaying local repo (editable)..."
|
Write-Host "==> Overlaying local repo (editable)..."
|
||||||
|
|
@ -709,7 +709,7 @@ shell.Run cmd, 0, False
|
||||||
if ($SkipTorch) {
|
if ($SkipTorch) {
|
||||||
# No-torch: install unsloth + unsloth-zoo with --no-deps, then
|
# No-torch: install unsloth + unsloth-zoo with --no-deps, then
|
||||||
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
||||||
uv pip install --python $VenvPython --no-deps --upgrade-package unsloth --upgrade-package unsloth-zoo "unsloth>=2026.3.14" unsloth-zoo
|
uv pip install --python $VenvPython --no-deps --upgrade-package unsloth --upgrade-package unsloth-zoo "unsloth>=2026.3.16" unsloth-zoo
|
||||||
$NoTorchReq = Find-NoTorchRuntimeFile
|
$NoTorchReq = Find-NoTorchRuntimeFile
|
||||||
if ($NoTorchReq) {
|
if ($NoTorchReq) {
|
||||||
uv pip install --python $VenvPython --no-deps -r $NoTorchReq
|
uv pip install --python $VenvPython --no-deps -r $NoTorchReq
|
||||||
|
|
@ -719,7 +719,7 @@ shell.Run cmd, 0, False
|
||||||
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
||||||
}
|
}
|
||||||
} elseif ($StudioLocalInstall) {
|
} elseif ($StudioLocalInstall) {
|
||||||
uv pip install --python $VenvPython --upgrade-package unsloth "unsloth>=2026.3.14" unsloth-zoo
|
uv pip install --python $VenvPython --upgrade-package unsloth "unsloth>=2026.3.16" unsloth-zoo
|
||||||
Write-Host "==> Overlaying local repo (editable)..."
|
Write-Host "==> Overlaying local repo (editable)..."
|
||||||
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
||||||
} else {
|
} else {
|
||||||
|
|
@ -729,7 +729,7 @@ shell.Run cmd, 0, False
|
||||||
# Fallback: GPU detection failed to produce a URL -- let uv resolve torch
|
# Fallback: GPU detection failed to produce a URL -- let uv resolve torch
|
||||||
Write-Host "==> Installing unsloth (this may take a few minutes)..."
|
Write-Host "==> Installing unsloth (this may take a few minutes)..."
|
||||||
if ($StudioLocalInstall) {
|
if ($StudioLocalInstall) {
|
||||||
uv pip install --python $VenvPython unsloth-zoo "unsloth>=2026.3.14" --torch-backend=auto
|
uv pip install --python $VenvPython unsloth-zoo "unsloth>=2026.3.16" --torch-backend=auto
|
||||||
Write-Host "==> Overlaying local repo (editable)..."
|
Write-Host "==> Overlaying local repo (editable)..."
|
||||||
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
uv pip install --python $VenvPython -e $RepoRoot --no-deps
|
||||||
} else {
|
} else {
|
||||||
|
|
|
||||||
10
install.sh
10
install.sh
|
|
@ -968,7 +968,7 @@ if [ "$_MIGRATED" = true ]; then
|
||||||
# to prevent transitive torch resolution.
|
# to prevent transitive torch resolution.
|
||||||
uv pip install --python "$_VENV_PY" --no-deps \
|
uv pip install --python "$_VENV_PY" --no-deps \
|
||||||
--reinstall-package unsloth --reinstall-package unsloth-zoo \
|
--reinstall-package unsloth --reinstall-package unsloth-zoo \
|
||||||
"unsloth>=2026.3.14" unsloth-zoo
|
"unsloth>=2026.3.16" unsloth-zoo
|
||||||
_NO_TORCH_RT="$(_find_no_torch_runtime)"
|
_NO_TORCH_RT="$(_find_no_torch_runtime)"
|
||||||
if [ -n "$_NO_TORCH_RT" ]; then
|
if [ -n "$_NO_TORCH_RT" ]; then
|
||||||
uv pip install --python "$_VENV_PY" --no-deps -r "$_NO_TORCH_RT"
|
uv pip install --python "$_VENV_PY" --no-deps -r "$_NO_TORCH_RT"
|
||||||
|
|
@ -976,7 +976,7 @@ if [ "$_MIGRATED" = true ]; then
|
||||||
else
|
else
|
||||||
uv pip install --python "$_VENV_PY" \
|
uv pip install --python "$_VENV_PY" \
|
||||||
--reinstall-package unsloth --reinstall-package unsloth-zoo \
|
--reinstall-package unsloth --reinstall-package unsloth-zoo \
|
||||||
"unsloth>=2026.3.14" unsloth-zoo
|
"unsloth>=2026.3.16" unsloth-zoo
|
||||||
fi
|
fi
|
||||||
if [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
if [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
||||||
echo "==> Overlaying local repo (editable)..."
|
echo "==> Overlaying local repo (editable)..."
|
||||||
|
|
@ -998,7 +998,7 @@ elif [ -n "$TORCH_INDEX_URL" ]; then
|
||||||
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
# runtime deps (typer, safetensors, transformers, etc.) with --no-deps.
|
||||||
uv pip install --python "$_VENV_PY" --no-deps \
|
uv pip install --python "$_VENV_PY" --no-deps \
|
||||||
--upgrade-package unsloth --upgrade-package unsloth-zoo \
|
--upgrade-package unsloth --upgrade-package unsloth-zoo \
|
||||||
"unsloth>=2026.3.14" unsloth-zoo
|
"unsloth>=2026.3.16" unsloth-zoo
|
||||||
_NO_TORCH_RT="$(_find_no_torch_runtime)"
|
_NO_TORCH_RT="$(_find_no_torch_runtime)"
|
||||||
if [ -n "$_NO_TORCH_RT" ]; then
|
if [ -n "$_NO_TORCH_RT" ]; then
|
||||||
uv pip install --python "$_VENV_PY" --no-deps -r "$_NO_TORCH_RT"
|
uv pip install --python "$_VENV_PY" --no-deps -r "$_NO_TORCH_RT"
|
||||||
|
|
@ -1009,7 +1009,7 @@ elif [ -n "$TORCH_INDEX_URL" ]; then
|
||||||
fi
|
fi
|
||||||
elif [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
elif [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
||||||
uv pip install --python "$_VENV_PY" \
|
uv pip install --python "$_VENV_PY" \
|
||||||
--upgrade-package unsloth "unsloth>=2026.3.14" unsloth-zoo
|
--upgrade-package unsloth "unsloth>=2026.3.16" unsloth-zoo
|
||||||
echo "==> Overlaying local repo (editable)..."
|
echo "==> Overlaying local repo (editable)..."
|
||||||
uv pip install --python "$_VENV_PY" -e "$_REPO_ROOT" --no-deps
|
uv pip install --python "$_VENV_PY" -e "$_REPO_ROOT" --no-deps
|
||||||
else
|
else
|
||||||
|
|
@ -1020,7 +1020,7 @@ else
|
||||||
# Fallback: GPU detection failed to produce a URL -- let uv resolve torch
|
# Fallback: GPU detection failed to produce a URL -- let uv resolve torch
|
||||||
echo "==> Installing unsloth (this may take a few minutes)..."
|
echo "==> Installing unsloth (this may take a few minutes)..."
|
||||||
if [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
if [ "$STUDIO_LOCAL_INSTALL" = true ]; then
|
||||||
uv pip install --python "$_VENV_PY" unsloth-zoo "unsloth>=2026.3.14" --torch-backend=auto
|
uv pip install --python "$_VENV_PY" unsloth-zoo "unsloth>=2026.3.16" --torch-backend=auto
|
||||||
echo "==> Overlaying local repo (editable)..."
|
echo "==> Overlaying local repo (editable)..."
|
||||||
uv pip install --python "$_VENV_PY" -e "$_REPO_ROOT" --no-deps
|
uv pip install --python "$_VENV_PY" -e "$_REPO_ROOT" --no-deps
|
||||||
else
|
else
|
||||||
|
|
|
||||||
636
pyproject.toml
636
pyproject.toml
|
|
@ -25,6 +25,29 @@ classifiers = [
|
||||||
"Topic :: Scientific/Engineering :: Artificial Intelligence",
|
"Topic :: Scientific/Engineering :: Artificial Intelligence",
|
||||||
]
|
]
|
||||||
dependencies = [
|
dependencies = [
|
||||||
|
"unsloth_zoo>=2026.3.6",
|
||||||
|
"wheel>=0.42.0",
|
||||||
|
"packaging",
|
||||||
|
"torch>=2.4.0,<2.11.0",
|
||||||
|
"torchvision",
|
||||||
|
"numpy",
|
||||||
|
"tqdm",
|
||||||
|
"psutil",
|
||||||
|
"tyro",
|
||||||
|
"protobuf",
|
||||||
|
"xformers>=0.0.27.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
|
"triton>=3.0.0 ; ('linux' in sys_platform)",
|
||||||
|
"triton-windows ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
|
"sentencepiece>=0.2.0",
|
||||||
|
"datasets>=3.4.1,!=4.0.*,!=4.1.0,<4.4.0",
|
||||||
|
"accelerate>=0.34.1",
|
||||||
|
"peft>=0.18.0,!=0.11.0",
|
||||||
|
"huggingface_hub>=0.34.0",
|
||||||
|
"hf_transfer",
|
||||||
|
"diffusers",
|
||||||
|
"transformers>=4.51.3,!=4.52.0,!=4.52.1,!=4.52.2,!=4.52.3,!=4.53.0,!=4.54.0,!=4.55.0,!=4.55.1,!=4.57.0,!=4.57.4,!=4.57.5,!=5.0.0,!=5.1.0,<=5.3.0",
|
||||||
|
"trl>=0.18.2,!=0.19.0,<=0.24.0",
|
||||||
"typer",
|
"typer",
|
||||||
"pydantic",
|
"pydantic",
|
||||||
"pyyaml",
|
"pyyaml",
|
||||||
|
|
@ -61,14 +84,11 @@ studio = [
|
||||||
exclude = ["images*", "tests*", "kernels/moe*"]
|
exclude = ["images*", "tests*", "kernels/moe*"]
|
||||||
|
|
||||||
[project.optional-dependencies]
|
[project.optional-dependencies]
|
||||||
triton = [
|
huggingface = [
|
||||||
"triton>=3.0.0 ; ('linux' in sys_platform)",
|
"unsloth_zoo>=2026.3.6",
|
||||||
"triton-windows ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
|
|
||||||
huggingfacenotorch = [
|
|
||||||
"wheel>=0.42.0",
|
"wheel>=0.42.0",
|
||||||
"packaging",
|
"packaging",
|
||||||
|
"torchvision",
|
||||||
"numpy",
|
"numpy",
|
||||||
"tqdm",
|
"tqdm",
|
||||||
"psutil",
|
"psutil",
|
||||||
|
|
@ -83,257 +103,132 @@ huggingfacenotorch = [
|
||||||
"diffusers",
|
"diffusers",
|
||||||
"transformers>=4.51.3,!=4.52.0,!=4.52.1,!=4.52.2,!=4.52.3,!=4.53.0,!=4.54.0,!=4.55.0,!=4.55.1,!=4.57.0,!=4.57.4,!=4.57.5,!=5.0.0,!=5.1.0,<=5.3.0",
|
"transformers>=4.51.3,!=4.52.0,!=4.52.1,!=4.52.2,!=4.52.3,!=4.53.0,!=4.54.0,!=4.55.0,!=4.55.1,!=4.57.0,!=4.57.4,!=4.57.5,!=5.0.0,!=5.1.0,<=5.3.0",
|
||||||
"trl>=0.18.2,!=0.19.0,<=0.24.0",
|
"trl>=0.18.2,!=0.19.0,<=0.24.0",
|
||||||
"sentence-transformers",
|
"typer",
|
||||||
]
|
"pydantic",
|
||||||
huggingface = [
|
"pyyaml",
|
||||||
"unsloth[huggingfacenotorch]",
|
"nest-asyncio",
|
||||||
"unsloth_zoo>=2026.3.6",
|
|
||||||
"torchvision",
|
|
||||||
"unsloth[triton]",
|
|
||||||
]
|
|
||||||
windows = [
|
|
||||||
"unsloth[huggingface]",
|
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0 ; (sys_platform == 'win32')",
|
|
||||||
"xformers>=0.0.22.post7 ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
base = [
|
base = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
]
|
]
|
||||||
cu118only = [
|
cu118only = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.22.post7%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.22.post7 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.22.post7%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.22.post7%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121only = [
|
cu121only = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.22.post7-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.22.post7 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.22.post7-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.22.post7-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch211 = [
|
cu118onlytorch211 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.23 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch211 = [
|
cu121onlytorch211 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.23 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch212 = [
|
cu118onlytorch212 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23.post1%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.23.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23.post1%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.23.post1%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch212 = [
|
cu121onlytorch212 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23.post1-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.23.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23.post1-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.23.post1-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch220 = [
|
cu118onlytorch220 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.24%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.24 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.24%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.24%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch220 = [
|
cu121onlytorch220 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.24-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.24 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.24-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.24-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch230 = [
|
cu118onlytorch230 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.27 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27%2Bcu118-cp312-cp312-manylinux2014_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch230 = [
|
cu121onlytorch230 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.27-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.27 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.27-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.27-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.27-cp312-cp312-manylinux2014_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch240 = [
|
cu118onlytorch240 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27.post2%2Bcu118-cp39-cp39-manylinux2014_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.27.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27.post2%2Bcu118-cp310-cp310-manylinux2014_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27.post2%2Bcu118-cp311-cp311-manylinux2014_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.27.post2%2Bcu118-cp312-cp312-manylinux2014_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch240 = [
|
cu121onlytorch240 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post1-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.27.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post1-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post1-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post1-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
|
||||||
cu124onlytorch240 = [
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post1-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch250 = [
|
cu118onlytorch250 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.28.post2-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.28.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.28.post2-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.28.post2-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.28.post2-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch250 = [
|
cu121onlytorch250 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post2-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.28.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post2-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post2-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.28.post2-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu124onlytorch250 = [
|
cu124onlytorch250 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.28.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.28.post2-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch251 = [
|
cu118onlytorch251 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post1-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post1-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post1-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post1-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu121onlytorch251 = [
|
cu121onlytorch251 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.29.post1-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.29.post1-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.29.post1-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu121/xformers-0.0.29.post1-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu124onlytorch251 = [
|
cu124onlytorch251 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post1-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch260 = [
|
cu118onlytorch260 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post3-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post3 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post3-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post3-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.29.post3-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
cu124onlytorch260 = [
|
cu124onlytorch260 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post3 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu124/xformers-0.0.29.post3-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch260 = [
|
cu126onlytorch260 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.29.post3 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.29.post3-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch270 = [
|
cu118onlytorch270 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.30 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.30-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch270 = [
|
cu126onlytorch270 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.30 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.30-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch270 = [
|
cu128onlytorch270 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp39-cp39-manylinux_2_28_x86_64.whl ; python_version=='3.9' and ('linux' in sys_platform)",
|
"xformers==0.0.30 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp310-cp310-manylinux_2_28_x86_64.whl ; python_version=='3.10' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp311-cp311-manylinux_2_28_x86_64.whl ; python_version=='3.11' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp312-cp312-manylinux_2_28_x86_64.whl ; python_version=='3.12' and ('linux' in sys_platform)",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp39-cp39-win_amd64.whl ; python_version=='3.9' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp310-cp310-win_amd64.whl ; python_version=='3.10' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp311-cp311-win_amd64.whl ; python_version=='3.11' and (sys_platform == 'win32')",
|
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.30-cp312-cp312-win_amd64.whl ; python_version=='3.12' and (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch271 = [
|
cu118onlytorch271 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.31.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.31.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu118/xformers-0.0.31.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch271 = [
|
cu126onlytorch271 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.31.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.31.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.31.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch271 = [
|
cu128onlytorch271 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.31.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.31.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.31.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118onlytorch280 = [
|
cu118onlytorch280 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.32.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.32.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.32.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch280 = [
|
cu126onlytorch280 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.32.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.32.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.32.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch280 = [
|
cu128onlytorch280 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu129/xformers-0.0.32.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.32.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu129/xformers-0.0.32.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu130onlytorch280 = [
|
cu130onlytorch280 = [
|
||||||
]
|
]
|
||||||
cu126onlytorch290 = [
|
cu126onlytorch290 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.33.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.33.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch290 = [
|
cu128onlytorch290 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.33.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.33.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu130onlytorch290 = [
|
cu130onlytorch290 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.33.post1-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post1 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.33.post1-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch291 = [
|
cu126onlytorch291 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.33.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.33.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch291 = [
|
cu128onlytorch291 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.33.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.33.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu130onlytorch291 = [
|
cu130onlytorch291 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.33.post2-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.33.post2 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.33.post2-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu126onlytorch2100 = [
|
cu126onlytorch2100 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.34-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.34 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu126/xformers-0.0.34-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu128onlytorch2100 = [
|
cu128onlytorch2100 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.34-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.34 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu128/xformers-0.0.34-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu130onlytorch2100 = [
|
cu130onlytorch2100 = [
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.34-cp39-abi3-manylinux_2_28_x86_64.whl ; ('linux' in sys_platform)",
|
"xformers==0.0.34 ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
||||||
"xformers @ https://download.pytorch.org/whl/cu130/xformers-0.0.34-cp39-abi3-win_amd64.whl ; (sys_platform == 'win32')",
|
|
||||||
]
|
]
|
||||||
cu118 = [
|
cu118 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
|
|
@ -347,22 +242,22 @@ cu121 = [
|
||||||
]
|
]
|
||||||
cu118-torch211 = [
|
cu118-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu118onlytorch211]",
|
"unsloth[cu118onlytorch211]",
|
||||||
]
|
]
|
||||||
cu121-torch211 = [
|
cu121-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch211]",
|
"unsloth[cu121onlytorch211]",
|
||||||
]
|
]
|
||||||
cu118-torch212 = [
|
cu118-torch212 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu118onlytorch212]",
|
"unsloth[cu118onlytorch212]",
|
||||||
]
|
]
|
||||||
cu121-torch212 = [
|
cu121-torch212 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch212]",
|
"unsloth[cu121onlytorch212]",
|
||||||
]
|
]
|
||||||
cu118-torch220 = [
|
cu118-torch220 = [
|
||||||
|
|
@ -432,17 +327,17 @@ cu124-torch251 = [
|
||||||
]
|
]
|
||||||
cu118-torch260 = [
|
cu118-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu118onlytorch260]",
|
"unsloth[cu118onlytorch260]",
|
||||||
]
|
]
|
||||||
cu124-torch260 = [
|
cu124-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu124onlytorch260]",
|
"unsloth[cu124onlytorch260]",
|
||||||
]
|
]
|
||||||
cu126-torch260 = [
|
cu126-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu126onlytorch260]",
|
"unsloth[cu126onlytorch260]",
|
||||||
]
|
]
|
||||||
cu118-torch270 = [
|
cu118-torch270 = [
|
||||||
|
|
@ -552,16 +447,19 @@ conda = [
|
||||||
]
|
]
|
||||||
colab-torch211 = [
|
colab-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch211]",
|
"unsloth[cu121onlytorch211]",
|
||||||
]
|
]
|
||||||
|
flashattention = [
|
||||||
|
"packaging ; ('linux' in sys_platform)",
|
||||||
|
"ninja ; ('linux' in sys_platform)",
|
||||||
|
"flash-attn>=2.6.3 ; ('linux' in sys_platform)",
|
||||||
|
]
|
||||||
colab-ampere-torch211 = [
|
colab-ampere-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch211]",
|
"unsloth[cu121onlytorch211]",
|
||||||
"packaging",
|
"unsloth[flashattention]",
|
||||||
"ninja",
|
|
||||||
"flash-attn>=2.6.3 ; ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
colab-torch220 = [
|
colab-torch220 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
|
|
@ -572,9 +470,7 @@ colab-ampere-torch220 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch220]",
|
"unsloth[cu121onlytorch220]",
|
||||||
"packaging",
|
"unsloth[flashattention]",
|
||||||
"ninja",
|
|
||||||
"flash-attn>=2.6.3 ; ('linux' in sys_platform)",
|
|
||||||
]
|
]
|
||||||
colab-new = [
|
colab-new = [
|
||||||
"unsloth_zoo>=2026.3.6",
|
"unsloth_zoo>=2026.3.6",
|
||||||
|
|
@ -592,24 +488,22 @@ colab-new = [
|
||||||
"hf_transfer",
|
"hf_transfer",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[triton]",
|
"unsloth[triton]",
|
||||||
"sentence-transformers",
|
"typer",
|
||||||
|
"pydantic",
|
||||||
|
"pyyaml",
|
||||||
|
"nest-asyncio",
|
||||||
]
|
]
|
||||||
colab-no-deps = [
|
colab-no-deps = [
|
||||||
"accelerate>=0.34.1",
|
"accelerate>=0.34.1",
|
||||||
"trl>=0.18.2,!=0.19.0,<=0.24.0",
|
"trl>=0.18.2,!=0.19.0,<=0.24.0",
|
||||||
"peft>=0.18.0",
|
"peft>=0.18.0",
|
||||||
"xformers ; ('linux' in sys_platform or sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
"xformers<0.0.27",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"protobuf",
|
"protobuf",
|
||||||
]
|
]
|
||||||
colab = [
|
colab = [
|
||||||
"unsloth[cu121]",
|
"unsloth[cu121]",
|
||||||
]
|
]
|
||||||
flashattention = [
|
|
||||||
"packaging ; ('linux' in sys_platform)",
|
|
||||||
"ninja ; ('linux' in sys_platform)",
|
|
||||||
"flash-attn>=2.6.3 ; ('linux' in sys_platform)",
|
|
||||||
]
|
|
||||||
colab-ampere = [
|
colab-ampere = [
|
||||||
"unsloth[colab-ampere-torch220]",
|
"unsloth[colab-ampere-torch220]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
|
|
@ -628,13 +522,13 @@ cu121-ampere = [
|
||||||
]
|
]
|
||||||
cu118-ampere-torch211 = [
|
cu118-ampere-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu118onlytorch211]",
|
"unsloth[cu118onlytorch211]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
]
|
]
|
||||||
cu121-ampere-torch211 = [
|
cu121-ampere-torch211 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes==0.45.5",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu121onlytorch211]",
|
"unsloth[cu121onlytorch211]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
]
|
]
|
||||||
|
|
@ -718,19 +612,19 @@ cu124-ampere-torch251 = [
|
||||||
]
|
]
|
||||||
cu118-ampere-torch260 = [
|
cu118-ampere-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu118onlytorch260]",
|
"unsloth[cu118onlytorch260]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
]
|
]
|
||||||
cu124-ampere-torch260 = [
|
cu124-ampere-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu124onlytorch260]",
|
"unsloth[cu124onlytorch260]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
]
|
]
|
||||||
cu126-ampere-torch260 = [
|
cu126-ampere-torch260 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.1",
|
||||||
"unsloth[cu126onlytorch260]",
|
"unsloth[cu126onlytorch260]",
|
||||||
"unsloth[flashattention]",
|
"unsloth[flashattention]",
|
||||||
]
|
]
|
||||||
|
|
@ -792,7 +686,6 @@ cu130-ampere-torch280 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu130onlytorch280]",
|
"unsloth[cu130onlytorch280]",
|
||||||
"unsloth[flashattention]",
|
|
||||||
]
|
]
|
||||||
cu126-ampere-torch290 = [
|
cu126-ampere-torch290 = [
|
||||||
"unsloth[huggingface]",
|
"unsloth[huggingface]",
|
||||||
|
|
@ -839,347 +732,8 @@ cu130-ampere-torch2100 = [
|
||||||
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
"bitsandbytes>=0.45.5,!=0.46.0,!=0.48.0",
|
||||||
"unsloth[cu130onlytorch2100]",
|
"unsloth[cu130onlytorch2100]",
|
||||||
]
|
]
|
||||||
flashattentiontorch260abiFALSEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiFALSE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
flashattentiontorch260abiTRUEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiTRUE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiTRUE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiTRUE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiTRUE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.6cxx11abiTRUE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
flashattentiontorch250abiFALSEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiFALSE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiFALSE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
flashattentiontorch250abiTRUEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiTRUE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiTRUE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiTRUE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiTRUE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.5cxx11abiTRUE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
flashattentiontorch240abiFALSEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiFALSE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiFALSE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiFALSE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiFALSE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiFALSE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
flashattentiontorch240abiTRUEcu12x = [
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiTRUE-cp39-cp39-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.9'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiTRUE-cp310-cp310-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.10'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiTRUE-cp311-cp311-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.11'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiTRUE-cp312-cp312-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.12'",
|
|
||||||
"flash-attn @ https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.4.post1/flash_attn-2.7.4.post1+cu12torch2.4cxx11abiTRUE-cp313-cp313-linux_x86_64.whl ; ('linux' in sys_platform) and python_version == '3.13'",
|
|
||||||
]
|
|
||||||
intelgputorch260 = [
|
|
||||||
"unsloth_zoo[intelgpu]",
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.2.0-cp39-cp39-linux_x86_64.whl#sha256=147607f190a7d7aa24ba454def5977fbbfec792fdae18e4ed278cfec29b69271 ; ('linux' in sys_platform) and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.2.0-cp310-cp310-linux_x86_64.whl#sha256=23aa423fa1542afc34f67eb3ba8ef20060f6d1b3a4697eaeab22b11c92b30f2b ; ('linux' in sys_platform) and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.2.0-cp311-cp311-linux_x86_64.whl#sha256=bcfa995229bbfd9ffd8d6c8d9f6428d393e876fa6e23ee3c20e3c0d73ca75ca5 ; ('linux' in sys_platform) and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.2.0-cp312-cp312-linux_x86_64.whl#sha256=bd340903d03470708df3442438acb8b7e08087ab9e61fbe349b2872bf9257ab0 ; ('linux' in sys_platform) and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.2.0-cp313-cp313-linux_x86_64.whl#sha256=814dccc8a07159e6eca74bed70091bc8fea2d9dd87b0d91845f9f38cde62f01c ; ('linux' in sys_platform) and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-manylinux_2_24_x86_64.whl ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-win_amd64.whl ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.6.0%2Bxpu-cp39-cp39-linux_x86_64.whl#sha256=6a8adf6dc4c089406e8b3a7e58ab57a463bddf9b07130d2576e76eced43e92af ; ('linux' in sys_platform) and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.6.0%2Bxpu-cp310-cp310-linux_x86_64.whl#sha256=ff4561cbf07c83bbccaa0f6e9bb0e6dcf721bacd53c9c43c4eb0e7331b4792f9 ; ('linux' in sys_platform) and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.6.0%2Bxpu-cp311-cp311-linux_x86_64.whl#sha256=12005f66b810ddd3ab93f86c4522bcfdd412cbd27fc9d189b661ff7509bc5e8a ; ('linux' in sys_platform) and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.6.0%2Bxpu-cp312-cp312-linux_x86_64.whl#sha256=c4c5c67625cdacf35765c2b94e61fe166e3c3f4a14521b1212a59ad1b3eb0f2e ; ('linux' in sys_platform) and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.6.0%2Bxpu-cp313-cp313-linux_x86_64.whl#sha256=e6864f7a60a5ecc43d5d38f59a16e5dd132384f73dfd3a697f74944026038f7b ; ('linux' in sys_platform) and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
intel-gpu-torch260 = [
|
|
||||||
"unsloth[intelgputorch260]"
|
|
||||||
]
|
|
||||||
intelgputorch270 = [
|
|
||||||
"unsloth_zoo[intelgpu]",
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp39-cp39-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=749a7098492c6a27b356c97149a4a62973b953eae60bc1b6259260974f344913 ; ('linux' in sys_platform) and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=44362e80abd752471a08341093321955b066daa2cfb4810e73b8e3b240850f93 ; ('linux' in sys_platform) and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=faa6b8c945a837a080f641bc8ccc77a98fa66980dcd7e62e715fd853737343fd ; ('linux' in sys_platform) and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=40f6fb65b345dc9a61813abe7ac9a585f2c9808f414d140cc2a5f11f53ee063c ; ('linux' in sys_platform) and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=b22b4c02ec71b4bfc862ae3cdfd2871dc0b05d2b1802f5db2196e0f897d581e9 ; ('linux' in sys_platform) and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp39-cp39-win_amd64.whl#sha256=d4b738d7fa5100c1bd766f91614962828a4810eb57b4df92cd5214a83505a752 ; sys_platform == 'win32' and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp310-cp310-win_amd64.whl#sha256=143fe8a64d807bcdb7d81bbc062816add325570aa160448454ab6ded4a0a17a1 ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp311-cp311-win_amd64.whl#sha256=a8025459ff325d6e3532eb5cf72519db1b178155e7d60aff6c56beb5968fc758 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp312-cp312-win_amd64.whl#sha256=0dd07e6d5b872e42e48f5ee140e609d4554ca3cc509d5bf509ac232267cf358e ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.3.0-cp313-cp313-win_amd64.whl#sha256=a936a18182d8e065a9933afc9a3ebbffadd38604969f87c493831214539fc027 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-manylinux_2_24_x86_64.whl ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-win_amd64.whl ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp39-cp39-linux_x86_64.whl#sha256=f8ee75e50fcbb37ed5b498299ca2264da99ab278a93fae2358e921e4a6e28273 ; ('linux' in sys_platform) and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp310-cp310-linux_x86_64.whl#sha256=d6fdc342961d98fdcd9d03dfd491a3208bb5f7fbb435841f8f72ce9fdcd2d026 ; ('linux' in sys_platform) and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp311-cp311-linux_x86_64.whl#sha256=74d07f9357df5cf2bf223ad3c84de16346bfaa0504f988fdd5590d3e177e5e86 ; ('linux' in sys_platform) and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp312-cp312-linux_x86_64.whl#sha256=c806d44aa2ca5d225629f6fbc6c994d5deaac2d2cde449195bc8e3522ddd219a ; ('linux' in sys_platform) and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp313-cp313-linux_x86_64.whl#sha256=25d8277b7f01d42e2e014ccbab57a2692b6ec4eff8dcf894eda1b297407cf97a ; ('linux' in sys_platform) and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp39-cp39-win_amd64.whl#sha256=046e85125266ae69c1a0d083e6c092f947ab4b6b41532c16bafe40dbced845df ; sys_platform == 'win32' and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=9ebaeffb82b0b3e39b6030927d3ebe0eb62a0e9045a3b2d7b0a9e7b15222c0db ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=356ba66cee127e7e2c942880bd50e03768306a4ea08d358a0f29c6eebfc4bc81 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=94739e665d9b4d5cd7af5f517cb6103f6f9fb421c095184609653a24524040f5 ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.7.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=31df3cb674918e89bc8c532baa331dc84f4430e1f9c0ec379232db44cba78355 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
intel-gpu-torch270 = [
|
|
||||||
"unsloth[intelgputorch270]"
|
|
||||||
]
|
|
||||||
intelgputorch280 = [
|
|
||||||
"unsloth_zoo[intelgpu]",
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp39-cp39-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=ac4d8e33986b1c3c5e48151640539272b2187e83016985853111b46fb82c3c94 ; 'linux' in sys_platform and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=999fef4c1f711092b9d3086525920545df490de476ecebe899ffc777019ae17f ; 'linux' in sys_platform and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=57b09c8c492985ff6a27cd3a22b08e8f7b96b407bd8030967b6efbb9f63b80cf ; 'linux' in sys_platform and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=df4bb3282bac9a3b90231700077110d8680b338416de03c2b7c6133c9b602649 ; 'linux' in sys_platform and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=60da63c99ca827bdcb0df28e0298bf7d066dc607454c6d6176783cb4e79d838b ; 'linux' in sys_platform and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp39-cp39-win_amd64.whl#sha256=64aea8de349f3e2e0ebf4c24b011a8122531fdffda5776edaef45829cc241cf8 ; sys_platform == 'win32' and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp310-cp310-win_amd64.whl#sha256=ae573d255b257fdbed319a3440dc9d0a721e31160ab7f6eba1b2226e6a409a1d ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp311-cp311-win_amd64.whl#sha256=8e0ea4558e5776d8ddab0264310be9b26aee5641bcac0da023537556d4317b86 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp312-cp312-win_amd64.whl#sha256=4090dde07a4fffc34aaf855701a9db28e9fccb57b368ade520f1a0f8e811c878 ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.4.0-cp313-cp313-win_amd64.whl#sha256=a33d0888f3c8df028a2d028842715837d0049524d6c06b9bb11869890a13601a ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp39-cp39-linux_x86_64.whl ; 'linux' in sys_platform and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp310-cp310-linux_x86_64.whl ; 'linux' in sys_platform and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp311-cp311-linux_x86_64.whl ; 'linux' in sys_platform and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp312-cp312-linux_x86_64.whl ; 'linux' in sys_platform and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp313-cp313-linux_x86_64.whl ; 'linux' in sys_platform and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp39-cp39-win_amd64.whl#sha256=f2f401276892428e4875cf1d8717c5cbab704b16fc594ccf23795e7b16549a99 ; sys_platform == 'win32' and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=125c60cd59d51b39581a7e9afcd4679bc3a6b8c1f9440b1bb502a23fdd60571e ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=47f1a57258cd460e80b38b2ed6744e31587ab77a96b4215bf59546cb4bab5cc0 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=0937d8943c145a83d9bafc6f80ef28971167817f9eda26066d33f72caf8a6646 ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.8.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=e034aab1d71760dc80a731531be43673ffe15e99033b82d24e40d2e6d41bd8bf ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-manylinux_2_24_x86_64.whl ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-win_amd64.whl ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp39-cp39-manylinux_2_28_x86_64.whl#sha256=6e981c192045fc249c008441179ff237bb00174d818b875b0475730b63f0eaca ; 'linux' in sys_platform and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp310-cp310-manylinux_2_28_x86_64.whl#sha256=e5ba4805969277175ebfd59cc717093528cc6e3ada89ac2725fc7a3c1fee6169 ; 'linux' in sys_platform and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp311-cp311-manylinux_2_28_x86_64.whl#sha256=74c39c144104416bc4c5ad8c26ab0c169dc5cc6be58059e01bc3665dd0ef676f ; 'linux' in sys_platform and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp312-cp312-manylinux_2_28_x86_64.whl#sha256=0acec355b80c3899841184084f365df336c508602812e34a44007b8b60d53af4 ; 'linux' in sys_platform and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp313-cp313-manylinux_2_28_x86_64.whl#sha256=e2109ae773dad27b98ca17681044b4f876563c37f2382b75de3a371399edcff8 ; 'linux' in sys_platform and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp39-cp39-win_amd64.whl#sha256=5f7904e7048d414379bc8c1167260f1e84204f105db2d0a2f9c89e87ce1cf205 ; sys_platform == 'win32' and python_version == '3.9' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=005fca5e658ca8e37adb63c1a021c84f5e56dfa6cf0d601d89cfe40b9473f79f ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=c6d030f5361461550c0ff1339b5bca8585fc1e84fda2e64b6184e65a581e4f98 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=91aafd61864cdce27461cbec13ddbf28c1bc6494265a1e4b80131c64a3b7d18f ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.23.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=71dc4a6421742ed1e7f585b04a100ad53615c341fbccfbc255aefb38ea9091da ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
intel-gpu-torch280 = [
|
|
||||||
"unsloth[intelgputorch280]"
|
|
||||||
]
|
|
||||||
intelgputorch290 = [
|
|
||||||
"unsloth_zoo[intelgpu]",
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=c169a1de14c19673b17c751290d467fa282fc90fa5da4314b2e5cdab1f553146 ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=013d9dd5d6479bd22983161f462e61c8dbe1d82e6730624a7a8d5945507eaa61 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=afc8cabfbf7ed51fd278d1e0f88d6afc157b0201bad4b99d681e4d542f9e66d4 ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=0d24c1716088f2764d0d24c64227732195b6a42706c3c5fc89eeb4904bfa0818 ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp310-cp310-win_amd64.whl#sha256=c83ab007311d9cfb6e809ee5a4587d99a9eef4be720b90da4f1aaa68b45139a0 ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp311-cp311-win_amd64.whl#sha256=debf75348da8e8c7166b4d4a9b91d1508bb8d6581e339f79f7604b2e6746bacd ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp312-cp312-win_amd64.whl#sha256=97337a47425f1963a723475bd61037460e84ba01db4f87a1d662c3718ff6c47e ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp313-cp313-win_amd64.whl#sha256=2caf8138695f6abb023ecd02031a2611ba1bf8fff2f19802567cb2fadefe9e87 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp310-cp310-linux_x86_64.whl#sha256=5afbe860ce991825a36b75706a523601087e414b77598ef0d9d3d565741c277d ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp311-cp311-linux_x86_64.whl#sha256=607fe419c32d6e8e0556f745742e7cff1d0babce51f54be890e0c1422359c442 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp312-cp312-linux_x86_64.whl#sha256=376bae584d89980b8e59934d248c38d5fa3b7d4687a4df1a19f4bc1d23dcc8c1 ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp313-cp313-linux_x86_64.whl#sha256=98d6a06dd7fb185874367b18bd609f05f16fdce4142a5980ca94461949965cd2 ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=47cc68f631f65bd9c84924d052cd04dec7531023caa85e80345e9c94611c887d ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=d56c44ab4818aba57e5c7b628f422d014e0d507427170a771c5be85e308b0bc6 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=18cad93aaff76a01ce73aef6935ece7cfc03344b905592ec731446c44d44592b ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.9.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=579929cdc10a76800ead41289cac191ea36d1b16f5f501d3fc25607d4375cd83 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-manylinux_2_24_x86_64.whl ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-win_amd64.whl ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp310-cp310-manylinux_2_28_x86_64.whl#sha256=cbfae2b79b7549fd368c2462fc8e94f8f26cc450782ee72138e908077c09a519 ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp311-cp311-manylinux_2_28_x86_64.whl#sha256=044fa36ef4b6b43edcd490b75c853fa4b3eb033c2bded29f8fbcf27734713c67 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp312-cp312-manylinux_2_28_x86_64.whl#sha256=4b91e4bec1d740a6211f02578a79888550b73f3a4e1383035f8f6d72f587212c ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp313-cp313-manylinux_2_28_x86_64.whl#sha256=88239e73ca37254bec84f29cd5887e10ff712de7edbbda3fbb3609cd6190d99e ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=19c7da8ca767d593e13a88a12bb08d06e34a673f6f26c2f9c191d60e81c02953 ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=9bb0d1421c544ac8e2eca5b47daacaf54706dc9139c003aa5e77ee5f355c5931 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=6a5194bc736089606342d48a3f6822829b167617e9495d91d753dd1bd46fda18 ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.24.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=da47a3ce2bb7f0301a31124668b5908f9b9e92d6241443de15a310ef9632fd83 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
intel-gpu-torch290 = [
|
|
||||||
"unsloth[intelgputorch290]"
|
|
||||||
]
|
|
||||||
intelgputorch210 = [
|
|
||||||
"unsloth_zoo[intelgpu]",
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=c169a1de14c19673b17c751290d467fa282fc90fa5da4314b2e5cdab1f553146 ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=013d9dd5d6479bd22983161f462e61c8dbe1d82e6730624a7a8d5945507eaa61 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=afc8cabfbf7ed51fd278d1e0f88d6afc157b0201bad4b99d681e4d542f9e66d4 ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl#sha256=0d24c1716088f2764d0d24c64227732195b6a42706c3c5fc89eeb4904bfa0818 ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp310-cp310-win_amd64.whl#sha256=c83ab007311d9cfb6e809ee5a4587d99a9eef4be720b90da4f1aaa68b45139a0 ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp311-cp311-win_amd64.whl#sha256=debf75348da8e8c7166b4d4a9b91d1508bb8d6581e339f79f7604b2e6746bacd ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp312-cp312-win_amd64.whl#sha256=97337a47425f1963a723475bd61037460e84ba01db4f87a1d662c3718ff6c47e ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"pytorch_triton_xpu @ https://download.pytorch.org/whl/pytorch_triton_xpu-3.5.0-cp313-cp313-win_amd64.whl#sha256=2caf8138695f6abb023ecd02031a2611ba1bf8fff2f19802567cb2fadefe9e87 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp310-cp310-linux_x86_64.whl#sha256=abb1d1ec1ac672bac0ff35420c965f2df0c636ef9d94e2a830e34578489d0a57 ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp311-cp311-linux_x86_64.whl#sha256=71ad2f82da0f41eaec159f39fc85854e27c2391efa91b373e550648a6f4aaad3 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp312-cp312-linux_x86_64.whl#sha256=b473571d478912f92881cc13f15fa18f8463fb0fb8a068c96ed47a7d45a4da0a ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp313-cp313-linux_x86_64.whl#sha256=3bc64a746ff25a93de140902c60c9e819d7413f5cea1e88d80999c27a5901e9c ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=ce50691ab3fb6301d9b7bb8b3834cf5fa7152a2b5f91fd24c5efdc601a25b780 ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=cb9d37f21cb9fb7df67d62863f021c3144e8d8832b9ea8e8523ac308bc620ea1 ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=3ad605be4728b6d3a28a44d07dd794b1a9e45551b0057815bf25eb2a6d6a56a7 ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torch @ https://download.pytorch.org/whl/xpu/torch-2.10.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=2b4b56dd6c792aef82006904fa888692e3782e4ae5da27526801bad4898f05a5 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-manylinux_2_24_x86_64.whl ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"bitsandbytes @ https://github.com/bitsandbytes-foundation/bitsandbytes/releases/download/continuous-release_main/bitsandbytes-1.33.7.preview-py3-none-win_amd64.whl ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp310-cp310-manylinux_2_28_x86_64.whl#sha256=7e1e7b170fcf7161c8499b67156c5a05462243626dc0974010791a0bab4378d3 ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp311-cp311-manylinux_2_28_x86_64.whl#sha256=bd6add201bd7628af70437292e1447abb368e0b5f4ff9abd334ae435efd44792 ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp312-cp312-manylinux_2_28_x86_64.whl#sha256=6ad2543496bc29e59d3dd614a94d09aa9870318aedb66045344fffddfedd2cf8 ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp313-cp313-manylinux_2_28_x86_64.whl#sha256=80269f37865fcd8b57f20e4786efae2200bfa2b2727926c3c7acc82f0e7d3548 ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp310-cp310-win_amd64.whl#sha256=6b9485ba85dcba4d196d6134d9c3332fb228fb2556416bf0450a64e8a472fcba ; sys_platform == 'win32' and python_version == '3.10' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp311-cp311-win_amd64.whl#sha256=36cbaedf10f6412af5c89afd9aeea474e6a56a0050348ada8fabe1ecaf6b879e ; sys_platform == 'win32' and python_version == '3.11' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp312-cp312-win_amd64.whl#sha256=738357d97468d75fe3d510ac37e65130f2787f81d9bbc1518898f7396dc3403f ; sys_platform == 'win32' and python_version == '3.12' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
"torchvision @ https://download.pytorch.org/whl/xpu/torchvision-0.25.0%2Bxpu-cp313-cp313-win_amd64.whl#sha256=1c4b44b36a557f7381e3076fb8843366742238648441d607c8d049c6da0f8886 ; sys_platform == 'win32' and python_version == '3.13' and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
intel-gpu-torch210 = [
|
|
||||||
"unsloth[intelgputorch210]"
|
|
||||||
]
|
|
||||||
intel = [
|
|
||||||
"unsloth[intelgputorch280]",
|
|
||||||
]
|
|
||||||
amd = [
|
|
||||||
"unsloth[huggingfacenotorch]",
|
|
||||||
"bitsandbytes>=0.49.1 ; ('linux' in sys_platform) and (platform_machine == 'AMD64' or platform_machine == 'x86_64' or platform_machine == 'aarch64')",
|
|
||||||
"bitsandbytes>=0.49.1 ; (sys_platform == 'win32') and (platform_machine == 'AMD64' or platform_machine == 'x86_64')",
|
|
||||||
]
|
|
||||||
rocm702-torch280 = [
|
|
||||||
"unsloth[amd]",
|
|
||||||
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/triton-3.4.0%2Brocm7.0.2.gitf9e5bf54-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/triton-3.4.0%2Brocm7.0.2.gitf9e5bf54-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/triton-3.4.0%2Brocm7.0.2.gitf9e5bf54-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torch-2.8.0%2Brocm7.0.2.lw.git245bf6ed-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torch-2.8.0%2Brocm7.0.2.lw.git245bf6ed-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torch-2.8.0%2Brocm7.0.2.lw.git245bf6ed-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torchvision-0.23.0%2Brocm7.0.2.git824e8c87-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torchvision-0.23.0%2Brocm7.0.2.git824e8c87-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.0.2/torchvision-0.23.0%2Brocm7.0.2.git824e8c87-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
]
|
|
||||||
rocm72-torch291 = [
|
|
||||||
"unsloth[amd]",
|
|
||||||
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.5.1%2Brocm7.2.0.gita272dfa8-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.5.1%2Brocm7.2.0.gita272dfa8-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.5.1%2Brocm7.2.0.gita272dfa8-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.5.1%2Brocm7.2.0.gita272dfa8-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.9.1%2Brocm7.2.0.lw.git7e1940d4-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.9.1%2Brocm7.2.0.lw.git7e1940d4-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.9.1%2Brocm7.2.0.lw.git7e1940d4-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.9.1%2Brocm7.2.0.lw.git7e1940d4-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/windows/rocm-rel-7.2/torch-2.9.1%2Brocmsdk20260116-cp312-cp312-win_amd64.whl ; sys_platform == 'win32' and python_version == '3.12'",
|
|
||||||
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.24.0%2Brocm7.2.0.gitb919bd0c-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.24.0%2Brocm7.2.0.gitb919bd0c-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.24.0%2Brocm7.2.0.gitb919bd0c-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.24.0%2Brocm7.2.0.gitb919bd0c-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/windows/rocm-rel-7.2/torchvision-0.24.1%2Brocmsdk20260116-cp312-cp312-win_amd64.whl ; sys_platform == 'win32' and python_version == '3.12'",
|
|
||||||
]
|
|
||||||
rocm711-torch291 = [
|
|
||||||
"unsloth[amd]",
|
|
||||||
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.5.1%2Brocm7.1.1.gita272dfa8-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.5.1%2Brocm7.1.1.gita272dfa8-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.5.1%2Brocm7.1.1.gita272dfa8-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.5.1%2Brocm7.1.1.gita272dfa8-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.9.1%2Brocm7.1.1.lw.git351ff442-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.9.1%2Brocm7.1.1.lw.git351ff442-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.9.1%2Brocm7.1.1.lw.git351ff442-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.9.1%2Brocm7.1.1.lw.git351ff442-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.24.0%2Brocm7.1.1.gitb919bd0c-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.24.0%2Brocm7.1.1.gitb919bd0c-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.24.0%2Brocm7.1.1.gitb919bd0c-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.24.0%2Brocm7.1.1.gitb919bd0c-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
]
|
|
||||||
rocm72-torch2100 = [
|
|
||||||
"unsloth[amd]",
|
|
||||||
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.6.0%2Brocm7.2.0.gitba5c1517-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.6.0%2Brocm7.2.0.gitba5c1517-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.6.0%2Brocm7.2.0.gitba5c1517-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/triton-3.6.0%2Brocm7.2.0.gitba5c1517-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.10.0%2Brocm7.2.0.lw.gitb6ee5fde-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.10.0%2Brocm7.2.0.lw.gitb6ee5fde-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.10.0%2Brocm7.2.0.lw.gitb6ee5fde-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torch-2.10.0%2Brocm7.2.0.lw.gitb6ee5fde-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.25.0%2Brocm7.2.0.git82df5f59-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.25.0%2Brocm7.2.0.git82df5f59-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.25.0%2Brocm7.2.0.git82df5f59-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.2/torchvision-0.25.0%2Brocm7.2.0.git82df5f59-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
]
|
|
||||||
rocm711-torch2100 = [
|
|
||||||
"unsloth[amd]",
|
|
||||||
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.6.0%2Brocm7.1.1.gitba5c1517-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.6.0%2Brocm7.1.1.gitba5c1517-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.6.0%2Brocm7.1.1.gitba5c1517-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"triton @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/triton-3.6.0%2Brocm7.1.1.gitba5c1517-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.10.0%2Brocm7.1.1.lw.gitd9556b05-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.10.0%2Brocm7.1.1.lw.gitd9556b05-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.10.0%2Brocm7.1.1.lw.gitd9556b05-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torch @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torch-2.10.0%2Brocm7.1.1.lw.gitd9556b05-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.25.0%2Brocm7.1.1.git82df5f59-cp310-cp310-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.10' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.25.0%2Brocm7.1.1.git82df5f59-cp311-cp311-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.11' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.25.0%2Brocm7.1.1.git82df5f59-cp312-cp312-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.12' and platform_machine == 'x86_64'",
|
|
||||||
"torchvision @ https://repo.radeon.com/rocm/manylinux/rocm-rel-7.1.1/torchvision-0.25.0%2Brocm7.1.1.git82df5f59-cp313-cp313-linux_x86_64.whl ; platform_system == 'Linux' and python_version == '3.13' and platform_machine == 'x86_64'",
|
|
||||||
]
|
|
||||||
|
|
||||||
[project.urls]
|
[project.urls]
|
||||||
homepage = "https://unsloth.ai"
|
homepage = "https://unsloth.ai"
|
||||||
documentation = "https://unsloth.ai/docs"
|
documentation = "https://unsloth.ai/docs"
|
||||||
repository = "https://github.com/unslothai/unsloth"
|
repository = "https://github.com/unslothai/unsloth"
|
||||||
|
|
||||||
[tool.ruff]
|
|
||||||
target-version = "py311"
|
|
||||||
force-exclude = true
|
|
||||||
extend-exclude = [
|
|
||||||
"*chat_templates.py",
|
|
||||||
"*ollama_template_mappers.py",
|
|
||||||
"*_auto_install.py",
|
|
||||||
"*mapper.py",
|
|
||||||
]
|
|
||||||
|
|
||||||
[tool.ruff.lint]
|
|
||||||
select = ["E9", "F63", "F7", "F82"]
|
|
||||||
ignore = [
|
|
||||||
"E402",
|
|
||||||
"E722",
|
|
||||||
"F403",
|
|
||||||
"F405",
|
|
||||||
"F811",
|
|
||||||
"F821",
|
|
||||||
"F841",
|
|
||||||
"F401",
|
|
||||||
"E731",
|
|
||||||
"E741",
|
|
||||||
"F601",
|
|
||||||
"E712",
|
|
||||||
]
|
|
||||||
|
|
||||||
[tool.ruff.format]
|
|
||||||
|
|
|
||||||
|
|
@ -82,7 +82,6 @@ def start(port: int = 8888):
|
||||||
from colab import start
|
from colab import start
|
||||||
start()
|
start()
|
||||||
"""
|
"""
|
||||||
import sys
|
|
||||||
|
|
||||||
logger.info("🦥 Starting Unsloth Studio...")
|
logger.info("🦥 Starting Unsloth Studio...")
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -4,8 +4,6 @@
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import structlog
|
|
||||||
import loggers
|
|
||||||
import logging
|
import logging
|
||||||
import re
|
import re
|
||||||
import shutil
|
import shutil
|
||||||
|
|
|
||||||
|
|
@ -5,7 +5,6 @@ from __future__ import annotations
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import os
|
import os
|
||||||
import structlog
|
|
||||||
import subprocess
|
import subprocess
|
||||||
from copy import deepcopy
|
from copy import deepcopy
|
||||||
from dataclasses import dataclass
|
from dataclasses import dataclass
|
||||||
|
|
|
||||||
|
|
@ -8,7 +8,6 @@ Export backend - handles model exporting in various formats
|
||||||
|
|
||||||
import glob
|
import glob
|
||||||
import json
|
import json
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import shutil
|
import shutil
|
||||||
|
|
@ -23,7 +22,7 @@ from utils.hardware import clear_gpu_cache
|
||||||
|
|
||||||
from utils.models import is_vision_model, get_base_model_from_lora
|
from utils.models import is_vision_model, get_base_model_from_lora
|
||||||
from utils.models.model_config import detect_audio_type
|
from utils.models.model_config import detect_audio_type
|
||||||
from utils.paths import ensure_dir, outputs_root, resolve_export_dir, resolve_output_dir
|
from utils.paths import ensure_dir, outputs_root, resolve_export_dir
|
||||||
from core.inference import get_inference_backend
|
from core.inference import get_inference_backend
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
|
||||||
|
|
@ -15,13 +15,11 @@ Pattern follows core/inference/orchestrator.py.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import atexit
|
import atexit
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import multiprocessing as mp
|
import multiprocessing as mp
|
||||||
import queue
|
import queue
|
||||||
import threading
|
import threading
|
||||||
import time
|
import time
|
||||||
from pathlib import Path
|
|
||||||
from typing import Any, List, Optional, Tuple
|
from typing import Any, List, Optional, Tuple
|
||||||
from utils.paths import outputs_root
|
from utils.paths import outputs_root
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -17,7 +17,6 @@ Pattern follows core/inference/worker.py and core/training/worker.py.
|
||||||
|
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import sys
|
import sys
|
||||||
|
|
|
||||||
|
|
@ -9,7 +9,6 @@ Supports: SNAC (Orpheus), CSM (Sesame), BiCodec (Spark), DAC (OuteTTS)
|
||||||
import io
|
import io
|
||||||
import re
|
import re
|
||||||
import wave
|
import wave
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from typing import Optional, Tuple
|
from typing import Optional, Tuple
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -7,21 +7,16 @@ Core inference backend - streamlined
|
||||||
|
|
||||||
from unsloth import FastLanguageModel, FastVisionModel
|
from unsloth import FastLanguageModel, FastVisionModel
|
||||||
from unsloth.chat_templates import get_chat_template
|
from unsloth.chat_templates import get_chat_template
|
||||||
from transformers import TextStreamer
|
|
||||||
from peft import PeftModel, PeftModelForCausalLM
|
from peft import PeftModel, PeftModelForCausalLM
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import sys
|
|
||||||
import torch
|
import torch
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import Optional, Union, Generator, Tuple
|
from typing import Optional, Union, Generator, Tuple
|
||||||
from utils.models import ModelConfig, get_base_model_from_lora
|
from utils.models import ModelConfig
|
||||||
from utils.paths import is_model_cached
|
|
||||||
from utils.utils import format_error_message
|
from utils.utils import format_error_message
|
||||||
from utils.hardware import get_device, clear_gpu_cache, log_gpu_memory
|
from utils.hardware import get_device, clear_gpu_cache, log_gpu_memory
|
||||||
from core.inference.audio_codecs import AudioCodecManager
|
from core.inference.audio_codecs import AudioCodecManager
|
||||||
from io import StringIO
|
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
|
|
||||||
|
|
@ -902,7 +897,6 @@ class InferenceBackend:
|
||||||
try:
|
try:
|
||||||
from utils.datasets import (
|
from utils.datasets import (
|
||||||
MODEL_TO_TEMPLATE_MAPPER,
|
MODEL_TO_TEMPLATE_MAPPER,
|
||||||
get_tokenizer_chat_template,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
model_name_lower = self.active_model_name.lower()
|
model_name_lower = self.active_model_name.lower()
|
||||||
|
|
@ -1143,7 +1137,6 @@ class InferenceBackend:
|
||||||
Uses processor.apply_chat_template with audio embedded in messages (Gemma 3n pattern).
|
Uses processor.apply_chat_template with audio embedded in messages (Gemma 3n pattern).
|
||||||
"""
|
"""
|
||||||
import threading
|
import threading
|
||||||
import numpy as np
|
|
||||||
|
|
||||||
model_info = self.models[self.active_model_name]
|
model_info = self.models[self.active_model_name]
|
||||||
model = model_info["model"]
|
model = model_info["model"]
|
||||||
|
|
@ -1736,7 +1729,7 @@ class InferenceBackend:
|
||||||
formatted_prompt = tokenizer.apply_chat_template(
|
formatted_prompt = tokenizer.apply_chat_template(
|
||||||
chat_messages, tokenize = False, add_generation_prompt = True
|
chat_messages, tokenize = False, add_generation_prompt = True
|
||||||
)
|
)
|
||||||
logger.info(f"Successfully applied tokenizer's native chat template")
|
logger.info("Successfully applied tokenizer's native chat template")
|
||||||
return formatted_prompt
|
return formatted_prompt
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
error_msg = str(e).lower()
|
error_msg = str(e).lower()
|
||||||
|
|
@ -1745,7 +1738,7 @@ class InferenceBackend:
|
||||||
or "no template argument" in error_msg
|
or "no template argument" in error_msg
|
||||||
):
|
):
|
||||||
logger.info(
|
logger.info(
|
||||||
f"Base model detected - no built-in chat template available, using fallback formatting"
|
"Base model detected - no built-in chat template available, using fallback formatting"
|
||||||
)
|
)
|
||||||
else:
|
else:
|
||||||
logger.warning(f"Failed to apply tokenizer chat template: {e}")
|
logger.warning(f"Failed to apply tokenizer chat template: {e}")
|
||||||
|
|
|
||||||
|
|
@ -12,7 +12,6 @@ import atexit
|
||||||
import contextlib
|
import contextlib
|
||||||
import json
|
import json
|
||||||
import struct
|
import struct
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import shutil
|
import shutil
|
||||||
import socket
|
import socket
|
||||||
|
|
@ -2034,7 +2033,7 @@ class LlamaCppBackend:
|
||||||
|
|
||||||
except httpx.ConnectError:
|
except httpx.ConnectError:
|
||||||
raise RuntimeError("Lost connection to llama-server")
|
raise RuntimeError("Lost connection to llama-server")
|
||||||
except Exception as e:
|
except Exception:
|
||||||
if cancel_event is not None and cancel_event.is_set():
|
if cancel_event is not None and cancel_event.is_set():
|
||||||
return
|
return
|
||||||
raise
|
raise
|
||||||
|
|
@ -2629,7 +2628,7 @@ class LlamaCppBackend:
|
||||||
|
|
||||||
except httpx.ConnectError:
|
except httpx.ConnectError:
|
||||||
raise RuntimeError("Lost connection to llama-server")
|
raise RuntimeError("Lost connection to llama-server")
|
||||||
except Exception as e:
|
except Exception:
|
||||||
if cancel_event is not None and cancel_event.is_set():
|
if cancel_event is not None and cancel_event.is_set():
|
||||||
return
|
return
|
||||||
raise
|
raise
|
||||||
|
|
@ -2792,7 +2791,7 @@ class LlamaCppBackend:
|
||||||
|
|
||||||
except httpx.ConnectError:
|
except httpx.ConnectError:
|
||||||
raise RuntimeError("Lost connection to llama-server")
|
raise RuntimeError("Lost connection to llama-server")
|
||||||
except Exception as e:
|
except Exception:
|
||||||
if cancel_event is not None and cancel_event.is_set():
|
if cancel_event is not None and cancel_event.is_set():
|
||||||
return
|
return
|
||||||
raise
|
raise
|
||||||
|
|
|
||||||
|
|
@ -17,7 +17,6 @@ Pattern follows core/training/training.py.
|
||||||
|
|
||||||
import atexit
|
import atexit
|
||||||
import base64
|
import base64
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import multiprocessing as mp
|
import multiprocessing as mp
|
||||||
import queue
|
import queue
|
||||||
|
|
@ -25,7 +24,6 @@ import threading
|
||||||
import time
|
import time
|
||||||
import uuid
|
import uuid
|
||||||
from io import BytesIO
|
from io import BytesIO
|
||||||
from pathlib import Path
|
|
||||||
from typing import Any, Generator, Optional, Tuple, Union
|
from typing import Any, Generator, Optional, Tuple, Union
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
|
||||||
|
|
@ -17,7 +17,6 @@ Pattern follows core/training/worker.py.
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import base64
|
import base64
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import queue as _queue
|
import queue as _queue
|
||||||
|
|
|
||||||
|
|
@ -33,22 +33,19 @@ if sys.platform in ("win32", "darwin"):
|
||||||
sys.path.insert(0, _compile_cache)
|
sys.path.insert(0, _compile_cache)
|
||||||
|
|
||||||
import torch
|
import torch
|
||||||
from utils.hardware import clear_gpu_cache, safe_num_proc, dataset_map_num_proc
|
from utils.hardware import clear_gpu_cache, dataset_map_num_proc
|
||||||
|
|
||||||
torch._dynamo.config.recompile_limit = 64
|
torch._dynamo.config.recompile_limit = 64
|
||||||
from unsloth import FastLanguageModel, FastVisionModel, is_bfloat16_supported
|
from unsloth import FastLanguageModel, FastVisionModel, is_bfloat16_supported
|
||||||
from unsloth.chat_templates import get_chat_template
|
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import threading
|
import threading
|
||||||
import math
|
import math
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import time
|
import time
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import Optional, Callable
|
from typing import Optional, Callable
|
||||||
from dataclasses import dataclass
|
from dataclasses import dataclass
|
||||||
import pandas as pd
|
|
||||||
from datasets import Dataset, load_dataset
|
from datasets import Dataset, load_dataset
|
||||||
|
|
||||||
from utils.models import is_vision_model, detect_audio_type
|
from utils.models import is_vision_model, detect_audio_type
|
||||||
|
|
@ -814,7 +811,7 @@ class UnslothTrainer:
|
||||||
# second attempt because the failed first call's partial
|
# second attempt because the failed first call's partial
|
||||||
# imports clean up the stale state as a side effect.
|
# imports clean up the stale state as a side effect.
|
||||||
self._source_code_retried = True
|
self._source_code_retried = True
|
||||||
logger.info(f"\n'could not get source code' — retrying once...\n")
|
logger.info("\n'could not get source code' — retrying once...\n")
|
||||||
return self.load_model(
|
return self.load_model(
|
||||||
model_name = model_name,
|
model_name = model_name,
|
||||||
max_seq_length = max_seq_length,
|
max_seq_length = max_seq_length,
|
||||||
|
|
@ -1016,7 +1013,7 @@ class UnslothTrainer:
|
||||||
# Phase 2: Whisper uses FastModel.get_peft_model with task_type=None
|
# Phase 2: Whisper uses FastModel.get_peft_model with task_type=None
|
||||||
from unsloth import FastModel
|
from unsloth import FastModel
|
||||||
|
|
||||||
logger.info(f"Audio model (whisper) LoRA configuration:")
|
logger.info("Audio model (whisper) LoRA configuration:")
|
||||||
logger.info(f" - Target modules: {target_modules}\n")
|
logger.info(f" - Target modules: {target_modules}\n")
|
||||||
|
|
||||||
self.model = FastModel.get_peft_model(
|
self.model = FastModel.get_peft_model(
|
||||||
|
|
@ -1057,7 +1054,7 @@ class UnslothTrainer:
|
||||||
|
|
||||||
elif self.is_vlm:
|
elif self.is_vlm:
|
||||||
# Vision model LoRA
|
# Vision model LoRA
|
||||||
logger.info(f"Vision model LoRA configuration:")
|
logger.info("Vision model LoRA configuration:")
|
||||||
logger.info(f" - Finetune vision layers: {finetune_vision_layers}")
|
logger.info(f" - Finetune vision layers: {finetune_vision_layers}")
|
||||||
logger.info(f" - Finetune language layers: {finetune_language_layers}")
|
logger.info(f" - Finetune language layers: {finetune_language_layers}")
|
||||||
logger.info(
|
logger.info(
|
||||||
|
|
@ -1085,7 +1082,7 @@ class UnslothTrainer:
|
||||||
)
|
)
|
||||||
else:
|
else:
|
||||||
# Text model LoRA
|
# Text model LoRA
|
||||||
logger.info(f"Text model LoRA configuration:")
|
logger.info("Text model LoRA configuration:")
|
||||||
logger.info(f" - Target modules: {target_modules}\n")
|
logger.info(f" - Target modules: {target_modules}\n")
|
||||||
|
|
||||||
self.model = FastLanguageModel.get_peft_model(
|
self.model = FastLanguageModel.get_peft_model(
|
||||||
|
|
@ -1114,7 +1111,6 @@ class UnslothTrainer:
|
||||||
|
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
import traceback
|
import traceback
|
||||||
import sys
|
|
||||||
|
|
||||||
error_details = (
|
error_details = (
|
||||||
f"{type(e).__name__}: {str(e)}"
|
f"{type(e).__name__}: {str(e)}"
|
||||||
|
|
@ -1140,7 +1136,6 @@ class UnslothTrainer:
|
||||||
and strip non-TransformersKwargs params that Unsloth/PEFT inject.
|
and strip non-TransformersKwargs params that Unsloth/PEFT inject.
|
||||||
"""
|
"""
|
||||||
import types
|
import types
|
||||||
import torch
|
|
||||||
import torch.nn as nn
|
import torch.nn as nn
|
||||||
from transformers.models.csm.modeling_csm import (
|
from transformers.models.csm.modeling_csm import (
|
||||||
CsmForConditionalGeneration,
|
CsmForConditionalGeneration,
|
||||||
|
|
@ -1707,7 +1702,6 @@ class UnslothTrainer:
|
||||||
"""
|
"""
|
||||||
import sys
|
import sys
|
||||||
import torch
|
import torch
|
||||||
import numpy as np
|
|
||||||
import torchaudio.transforms as T
|
import torchaudio.transforms as T
|
||||||
|
|
||||||
import subprocess
|
import subprocess
|
||||||
|
|
@ -2550,7 +2544,7 @@ class UnslothTrainer:
|
||||||
custom_format_mapping = custom_format_mapping,
|
custom_format_mapping = custom_format_mapping,
|
||||||
)
|
)
|
||||||
eval_dataset = eval_info["dataset"]
|
eval_dataset = eval_info["dataset"]
|
||||||
logger.info(f"Eval dataset formatted successfully\n")
|
logger.info("Eval dataset formatted successfully\n")
|
||||||
elif eval_enabled and not has_separate_eval_source:
|
elif eval_enabled and not has_separate_eval_source:
|
||||||
# No separate eval source — split the already-formatted dataset
|
# No separate eval source — split the already-formatted dataset
|
||||||
formatted_dataset = dataset_info["dataset"]
|
formatted_dataset = dataset_info["dataset"]
|
||||||
|
|
@ -3049,7 +3043,7 @@ class UnslothTrainer:
|
||||||
else:
|
else:
|
||||||
# Default to warmup_steps if neither provided
|
# Default to warmup_steps if neither provided
|
||||||
config_args["warmup_steps"] = 5
|
config_args["warmup_steps"] = 5
|
||||||
logger.info(f"Using default warmup_steps: 5\n")
|
logger.info("Using default warmup_steps: 5\n")
|
||||||
|
|
||||||
# Add save_steps if specified
|
# Add save_steps if specified
|
||||||
save_steps_val = training_args.get("save_steps", 0)
|
save_steps_val = training_args.get("save_steps", 0)
|
||||||
|
|
@ -3191,7 +3185,7 @@ class UnslothTrainer:
|
||||||
self.tokenizer, "tokenizer"
|
self.tokenizer, "tokenizer"
|
||||||
):
|
):
|
||||||
logger.info(
|
logger.info(
|
||||||
f" ⚠️ Unwrapping Processor → raw tokenizer for text-only SFTTrainer"
|
" ⚠️ Unwrapping Processor → raw tokenizer for text-only SFTTrainer"
|
||||||
)
|
)
|
||||||
sft_tokenizer = self.tokenizer.tokenizer
|
sft_tokenizer = self.tokenizer.tokenizer
|
||||||
|
|
||||||
|
|
@ -3499,7 +3493,6 @@ def _ensure_deepseek_ocr_installed():
|
||||||
sys.path.insert(0, parent_dir)
|
sys.path.insert(0, parent_dir)
|
||||||
|
|
||||||
# Try importing again
|
# Try importing again
|
||||||
from deepseek_ocr.modeling_deepseekocr import format_messages
|
|
||||||
|
|
||||||
logger.info("DeepSeek OCR module installed successfully")
|
logger.info("DeepSeek OCR module installed successfully")
|
||||||
logger.info("DeepSeek OCR module installed successfully!\n")
|
logger.info("DeepSeek OCR module installed successfully!\n")
|
||||||
|
|
|
||||||
|
|
@ -19,12 +19,9 @@ import math
|
||||||
import multiprocessing as mp
|
import multiprocessing as mp
|
||||||
import queue
|
import queue
|
||||||
import threading
|
import threading
|
||||||
import time
|
|
||||||
import structlog
|
|
||||||
from datetime import datetime, timezone
|
from datetime import datetime, timezone
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from dataclasses import dataclass, field
|
from dataclasses import dataclass
|
||||||
from pathlib import Path
|
|
||||||
from typing import Optional, Tuple, Any
|
from typing import Optional, Tuple, Any
|
||||||
|
|
||||||
import matplotlib.pyplot as plt
|
import matplotlib.pyplot as plt
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@ Pattern follows core/data_recipe/jobs/worker.py.
|
||||||
|
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import platform
|
import platform
|
||||||
|
|
@ -460,7 +459,6 @@ def run_training_process(
|
||||||
ensure_dir,
|
ensure_dir,
|
||||||
resolve_output_dir,
|
resolve_output_dir,
|
||||||
resolve_tensorboard_dir,
|
resolve_tensorboard_dir,
|
||||||
datasets_root,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
import transformers
|
import transformers
|
||||||
|
|
|
||||||
|
|
@ -37,7 +37,7 @@ if os.getenv("ENVIRONMENT_TYPE", "production") == "production":
|
||||||
from fastapi import Depends, FastAPI, Request
|
from fastapi import Depends, FastAPI, Request
|
||||||
from fastapi.middleware.cors import CORSMiddleware
|
from fastapi.middleware.cors import CORSMiddleware
|
||||||
from fastapi.staticfiles import StaticFiles
|
from fastapi.staticfiles import StaticFiles
|
||||||
from fastapi.responses import FileResponse, HTMLResponse, Response
|
from fastapi.responses import FileResponse, Response
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from datetime import datetime
|
from datetime import datetime
|
||||||
|
|
||||||
|
|
@ -219,7 +219,7 @@ async def get_system_info():
|
||||||
import platform
|
import platform
|
||||||
import subprocess
|
import subprocess
|
||||||
import psutil
|
import psutil
|
||||||
from utils.hardware import get_device, get_gpu_memory_info, DeviceType
|
from utils.hardware import get_gpu_memory_info
|
||||||
|
|
||||||
# GPU Info — query nvidia-smi for physical GPUs, filtered by
|
# GPU Info — query nvidia-smi for physical GPUs, filtered by
|
||||||
# CUDA_VISIBLE_DEVICES when set (the frontend uses this for GGUF
|
# CUDA_VISIBLE_DEVICES when set (the frontend uses this for GGUF
|
||||||
|
|
|
||||||
|
|
@ -6,7 +6,7 @@ Pydantic schemas for Export API.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
from pydantic import BaseModel, Field
|
from pydantic import BaseModel, Field
|
||||||
from typing import List, Optional, Literal, Dict, Any
|
from typing import Optional, Literal, Dict, Any
|
||||||
|
|
||||||
|
|
||||||
class LoadCheckpointRequest(BaseModel):
|
class LoadCheckpointRequest(BaseModel):
|
||||||
|
|
|
||||||
|
|
@ -18,7 +18,6 @@ from auth import storage, hashing
|
||||||
from auth.authentication import (
|
from auth.authentication import (
|
||||||
create_access_token,
|
create_access_token,
|
||||||
create_refresh_token,
|
create_refresh_token,
|
||||||
get_current_subject,
|
|
||||||
get_current_subject_allow_password_change,
|
get_current_subject_allow_password_change,
|
||||||
refresh_access_token,
|
refresh_access_token,
|
||||||
)
|
)
|
||||||
|
|
|
||||||
|
|
@ -12,7 +12,6 @@ import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from uuid import uuid4
|
from uuid import uuid4
|
||||||
from fastapi import APIRouter, Depends, HTTPException, UploadFile
|
from fastapi import APIRouter, Depends, HTTPException, UploadFile
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
# Add backend directory to path
|
# Add backend directory to path
|
||||||
|
|
|
||||||
|
|
@ -7,8 +7,7 @@ Export API routes: checkpoint discovery and model export operations.
|
||||||
|
|
||||||
import sys
|
import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from fastapi import APIRouter, Depends, HTTPException, Query
|
from fastapi import APIRouter, Depends, HTTPException
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
# Add backend directory to path
|
# Add backend directory to path
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@ from fastapi import APIRouter, Depends, HTTPException, Request
|
||||||
from fastapi.responses import StreamingResponse, JSONResponse
|
from fastapi.responses import StreamingResponse, JSONResponse
|
||||||
from typing import Optional
|
from typing import Optional
|
||||||
import json
|
import json
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import asyncio
|
import asyncio
|
||||||
import threading
|
import threading
|
||||||
|
|
@ -82,8 +81,6 @@ from models.inference import (
|
||||||
)
|
)
|
||||||
from auth.authentication import get_current_subject
|
from auth.authentication import get_current_subject
|
||||||
|
|
||||||
import io
|
|
||||||
import wave
|
|
||||||
import base64
|
import base64
|
||||||
import numpy as np
|
import numpy as np
|
||||||
|
|
||||||
|
|
@ -328,14 +325,14 @@ async def load_model(
|
||||||
training_method = adapter_cfg.get("unsloth_training_method")
|
training_method = adapter_cfg.get("unsloth_training_method")
|
||||||
if training_method == "lora" and load_in_4bit:
|
if training_method == "lora" and load_in_4bit:
|
||||||
logger.info(
|
logger.info(
|
||||||
f"adapter_config.json says unsloth_training_method='lora' — "
|
"adapter_config.json says unsloth_training_method='lora' — "
|
||||||
f"setting load_in_4bit=False to match 16-bit training"
|
"setting load_in_4bit=False to match 16-bit training"
|
||||||
)
|
)
|
||||||
load_in_4bit = False
|
load_in_4bit = False
|
||||||
elif training_method == "qlora" and not load_in_4bit:
|
elif training_method == "qlora" and not load_in_4bit:
|
||||||
logger.info(
|
logger.info(
|
||||||
f"adapter_config.json says unsloth_training_method='qlora' — "
|
"adapter_config.json says unsloth_training_method='qlora' — "
|
||||||
f"setting load_in_4bit=True to match QLoRA training"
|
"setting load_in_4bit=True to match QLoRA training"
|
||||||
)
|
)
|
||||||
load_in_4bit = True
|
load_in_4bit = True
|
||||||
elif training_method:
|
elif training_method:
|
||||||
|
|
@ -752,7 +749,6 @@ async def generate_audio(
|
||||||
|
|
||||||
def _decode_audio_base64(b64: str) -> np.ndarray:
|
def _decode_audio_base64(b64: str) -> np.ndarray:
|
||||||
"""Decode base64 audio (any format) → float32 numpy array at 16kHz."""
|
"""Decode base64 audio (any format) → float32 numpy array at 16kHz."""
|
||||||
import torch
|
|
||||||
import torchaudio
|
import torchaudio
|
||||||
import tempfile
|
import tempfile
|
||||||
import os
|
import os
|
||||||
|
|
|
||||||
|
|
@ -10,7 +10,6 @@ import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from fastapi import APIRouter, Body, Depends, HTTPException, Query
|
from fastapi import APIRouter, Body, Depends, HTTPException, Query
|
||||||
from typing import List, Optional
|
from typing import List, Optional
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
import re as _re
|
import re as _re
|
||||||
|
|
@ -775,7 +774,6 @@ async def get_gguf_variants(
|
||||||
# case-insensitive match.
|
# case-insensitive match.
|
||||||
cached_bytes_by_quant: dict[str, int] = {}
|
cached_bytes_by_quant: dict[str, int] = {}
|
||||||
try:
|
try:
|
||||||
import re as _re
|
|
||||||
from huggingface_hub import constants as hf_constants
|
from huggingface_hub import constants as hf_constants
|
||||||
|
|
||||||
# Sanitize repo_id: must be "owner/name" with safe chars only
|
# Sanitize repo_id: must be "owner/name" with safe chars only
|
||||||
|
|
|
||||||
|
|
@ -9,8 +9,7 @@ import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from fastapi import APIRouter, Depends, HTTPException, Request
|
from fastapi import APIRouter, Depends, HTTPException, Request
|
||||||
from fastapi.responses import StreamingResponse
|
from fastapi.responses import StreamingResponse
|
||||||
from typing import Dict, Optional, Any
|
from typing import Optional, Any
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import asyncio
|
import asyncio
|
||||||
from datetime import datetime
|
from datetime import datetime
|
||||||
|
|
|
||||||
|
|
@ -4,7 +4,6 @@
|
||||||
"""Tests for transformers version detection with local checkpoint fallbacks."""
|
"""Tests for transformers version detection with local checkpoint fallbacks."""
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import pytest
|
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from unittest.mock import patch
|
from unittest.mock import patch
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -17,7 +17,6 @@ Run with:
|
||||||
python -m pytest tests/test_utils.py -v
|
python -m pytest tests/test_utils.py -v
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import platform
|
|
||||||
from unittest.mock import patch, MagicMock
|
from unittest.mock import patch, MagicMock
|
||||||
|
|
||||||
import pytest
|
import pytest
|
||||||
|
|
@ -295,8 +294,6 @@ class TestLogGpuMemory:
|
||||||
"utilization_pct": 12.5,
|
"utilization_pct": 12.5,
|
||||||
"free_gb": 14.0,
|
"free_gb": 14.0,
|
||||||
}
|
}
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
|
||||||
|
|
||||||
with (
|
with (
|
||||||
patch(
|
patch(
|
||||||
|
|
@ -312,8 +309,6 @@ class TestLogGpuMemory:
|
||||||
|
|
||||||
def test_logs_cpu_fallback_when_no_gpu(self, caplog):
|
def test_logs_cpu_fallback_when_no_gpu(self, caplog):
|
||||||
fake_info = {"available": False, "backend": "cpu"}
|
fake_info = {"available": False, "backend": "cpu"}
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
|
||||||
|
|
||||||
with (
|
with (
|
||||||
patch(
|
patch(
|
||||||
|
|
|
||||||
|
|
@ -12,7 +12,6 @@ by spawned subprocesses.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import shutil
|
import shutil
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import List, Optional
|
from typing import List, Optional
|
||||||
|
|
|
||||||
|
|
@ -8,14 +8,13 @@ This module contains functions for applying chat templates to datasets
|
||||||
and generating dataset info summaries.
|
and generating dataset info summaries.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
from .format_detection import detect_dataset_format, detect_multimodal_dataset, detect_custom_format_heuristic
|
from .format_detection import detect_custom_format_heuristic
|
||||||
from .model_mappings import MODEL_TO_TEMPLATE_MAPPER
|
from .model_mappings import MODEL_TO_TEMPLATE_MAPPER
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
DEFAULT_ALPACA_TEMPLATE = """Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
|
DEFAULT_ALPACA_TEMPLATE = """Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
|
||||||
|
|
||||||
### Instruction:
|
### Instruction:
|
||||||
|
|
@ -63,18 +62,21 @@ def get_tokenizer_chat_template(tokenizer, model_name):
|
||||||
)
|
)
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
logger.info(f"⚠️ Failed to apply Unsloth template '{matched_template}': {e}")
|
logger.info(f"⚠️ Failed to apply Unsloth template '{matched_template}': {e}")
|
||||||
logger.info(f" Falling back to tokenizer's default chat template")
|
logger.info(" Falling back to tokenizer's default chat template")
|
||||||
else:
|
else:
|
||||||
# Check if tokenizer actually has a chat_template set
|
# Check if tokenizer actually has a chat_template set
|
||||||
has_chat_template = (
|
has_chat_template = (
|
||||||
hasattr(tokenizer, 'chat_template')
|
hasattr(tokenizer, "chat_template") and tokenizer.chat_template is not None
|
||||||
and tokenizer.chat_template is not None
|
|
||||||
)
|
)
|
||||||
if has_chat_template:
|
if has_chat_template:
|
||||||
logger.info(f"📝 Using tokenizer's own chat template (no Unsloth template match)")
|
logger.info(
|
||||||
|
"📝 Using tokenizer's own chat template (no Unsloth template match)"
|
||||||
|
)
|
||||||
else:
|
else:
|
||||||
# Base model with no chat template — apply default ChatML
|
# Base model with no chat template — apply default ChatML
|
||||||
logger.info(f"📝 No chat template found — applying default ChatML template (base model)")
|
logger.info(
|
||||||
|
"📝 No chat template found — applying default ChatML template (base model)"
|
||||||
|
)
|
||||||
try:
|
try:
|
||||||
tokenizer = get_chat_template(
|
tokenizer = get_chat_template(
|
||||||
tokenizer,
|
tokenizer,
|
||||||
|
|
@ -82,7 +84,7 @@ def get_tokenizer_chat_template(tokenizer, model_name):
|
||||||
)
|
)
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
logger.info(f"⚠️ Failed to apply default ChatML template: {e}")
|
logger.info(f"⚠️ Failed to apply default ChatML template: {e}")
|
||||||
logger.info(f" Falling back to tokenizer as-is")
|
logger.info(" Falling back to tokenizer as-is")
|
||||||
|
|
||||||
return tokenizer
|
return tokenizer
|
||||||
|
|
||||||
|
|
@ -99,7 +101,7 @@ def get_dataset_info_summary(dataset_info):
|
||||||
"sharegpt": "ShareGPT format (needs standardization)",
|
"sharegpt": "ShareGPT format (needs standardization)",
|
||||||
"chatml_messages": "ChatML format (messages column) - OpenAI compatible",
|
"chatml_messages": "ChatML format (messages column) - OpenAI compatible",
|
||||||
"chatml_conversations": "ChatML format (conversations column) - HuggingFace standard",
|
"chatml_conversations": "ChatML format (conversations column) - HuggingFace standard",
|
||||||
"unknown": "Unknown format"
|
"unknown": "Unknown format",
|
||||||
}
|
}
|
||||||
|
|
||||||
return {
|
return {
|
||||||
|
|
@ -110,7 +112,8 @@ def get_dataset_info_summary(dataset_info):
|
||||||
"chat_column": dataset_info["chat_column"],
|
"chat_column": dataset_info["chat_column"],
|
||||||
"is_standardized": dataset_info["is_standardized"],
|
"is_standardized": dataset_info["is_standardized"],
|
||||||
"warnings": dataset_info.get("warnings", []),
|
"warnings": dataset_info.get("warnings", []),
|
||||||
"ready_for_training": dataset_info["is_standardized"] and final_format != "unknown"
|
"ready_for_training": dataset_info["is_standardized"]
|
||||||
|
and final_format != "unknown",
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
||||||
|
|
@ -154,7 +157,7 @@ def apply_chat_template_to_dataset(
|
||||||
# Get EOS token if needed
|
# Get EOS token if needed
|
||||||
eos_token = ""
|
eos_token = ""
|
||||||
if add_eos_token:
|
if add_eos_token:
|
||||||
if hasattr(tokenizer, 'eos_token') and tokenizer.eos_token:
|
if hasattr(tokenizer, "eos_token") and tokenizer.eos_token:
|
||||||
eos_token = tokenizer.eos_token
|
eos_token = tokenizer.eos_token
|
||||||
else:
|
else:
|
||||||
warnings.append("add_eos_token=True but tokenizer has no eos_token")
|
warnings.append("add_eos_token=True but tokenizer has no eos_token")
|
||||||
|
|
@ -167,14 +170,16 @@ def apply_chat_template_to_dataset(
|
||||||
if not dataset_info.get("auto_detection_attempted", False):
|
if not dataset_info.get("auto_detection_attempted", False):
|
||||||
custom_format_mapping = detect_custom_format_heuristic(dataset)
|
custom_format_mapping = detect_custom_format_heuristic(dataset)
|
||||||
if custom_format_mapping:
|
if custom_format_mapping:
|
||||||
warnings.append(f"Auto-detected column mapping: {custom_format_mapping}")
|
warnings.append(
|
||||||
|
f"Auto-detected column mapping: {custom_format_mapping}"
|
||||||
|
)
|
||||||
else:
|
else:
|
||||||
errors.append("Could not auto-detect format mapping")
|
errors.append("Could not auto-detect format mapping")
|
||||||
return {
|
return {
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
else:
|
else:
|
||||||
# Already failed once in format_dataset, don't retry
|
# Already failed once in format_dataset, don't retry
|
||||||
|
|
@ -186,7 +191,7 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
|
|
||||||
if custom_format_mapping:
|
if custom_format_mapping:
|
||||||
|
|
@ -209,7 +214,7 @@ def apply_chat_template_to_dataset(
|
||||||
|
|
||||||
for i in range(num_examples):
|
for i in range(num_examples):
|
||||||
convo = []
|
convo = []
|
||||||
role_order = ['system', 'user', 'assistant']
|
role_order = ["system", "user", "assistant"]
|
||||||
|
|
||||||
for target_role in role_order:
|
for target_role in role_order:
|
||||||
for col_name, role in custom_format_mapping.items():
|
for col_name, role in custom_format_mapping.items():
|
||||||
|
|
@ -218,11 +223,18 @@ def apply_chat_template_to_dataset(
|
||||||
|
|
||||||
if is_user_provided:
|
if is_user_provided:
|
||||||
# User explicitly mapped - include even if empty
|
# User explicitly mapped - include even if empty
|
||||||
convo.append({"role": role, "content": str(content) if content else ""})
|
convo.append(
|
||||||
|
{
|
||||||
|
"role": role,
|
||||||
|
"content": str(content) if content else "",
|
||||||
|
}
|
||||||
|
)
|
||||||
else:
|
else:
|
||||||
# Auto-detected - skip empty
|
# Auto-detected - skip empty
|
||||||
if content and str(content).strip():
|
if content and str(content).strip():
|
||||||
convo.append({"role": role, "content": str(content)})
|
convo.append(
|
||||||
|
{"role": role, "content": str(content)}
|
||||||
|
)
|
||||||
|
|
||||||
conversations.append(convo)
|
conversations.append(convo)
|
||||||
|
|
||||||
|
|
@ -232,31 +244,35 @@ def apply_chat_template_to_dataset(
|
||||||
return result
|
return result
|
||||||
|
|
||||||
try:
|
try:
|
||||||
dataset = dataset.map(_apply_custom_mapping, batched = True, batch_size = batch_size)
|
dataset = dataset.map(
|
||||||
|
_apply_custom_mapping, batched = True, batch_size = batch_size
|
||||||
|
)
|
||||||
# Update to use conversations format
|
# Update to use conversations format
|
||||||
final_format = "chatml_conversations"
|
final_format = "chatml_conversations"
|
||||||
chat_column = "conversations"
|
chat_column = "conversations"
|
||||||
is_standardized = True
|
is_standardized = True
|
||||||
warnings.append("Successfully converted to ChatML format via custom mapping")
|
warnings.append(
|
||||||
|
"Successfully converted to ChatML format via custom mapping"
|
||||||
|
)
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
errors.append(f"Custom format mapping failed: {e}")
|
errors.append(f"Custom format mapping failed: {e}")
|
||||||
return {
|
return {
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
|
|
||||||
# ALPACA FORMAT
|
# ALPACA FORMAT
|
||||||
if final_format == "alpaca":
|
if final_format == "alpaca":
|
||||||
|
|
||||||
# Set alpaca chat template on tokenizer for saving (if not already set)
|
# Set alpaca chat template on tokenizer for saving (if not already set)
|
||||||
# This ensures the template is saved with the model for inference
|
# This ensures the template is saved with the model for inference
|
||||||
if not (hasattr(tokenizer, 'chat_template') and tokenizer.chat_template):
|
if not (hasattr(tokenizer, "chat_template") and tokenizer.chat_template):
|
||||||
try:
|
try:
|
||||||
from unsloth.chat_templates import get_chat_template
|
from unsloth.chat_templates import get_chat_template
|
||||||
|
|
||||||
tokenizer = get_chat_template(tokenizer, chat_template = "alpaca")
|
tokenizer = get_chat_template(tokenizer, chat_template = "alpaca")
|
||||||
logger.info(f"📝 Set alpaca chat template on tokenizer for model saving")
|
logger.info("📝 Set alpaca chat template on tokenizer for model saving")
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
logger.info(f"⚠️ Could not set alpaca template on tokenizer: {e}")
|
logger.info(f"⚠️ Could not set alpaca template on tokenizer: {e}")
|
||||||
|
|
||||||
|
|
@ -266,12 +282,16 @@ def apply_chat_template_to_dataset(
|
||||||
for i in range(len(examples["instruction"])):
|
for i in range(len(examples["instruction"])):
|
||||||
fields = {
|
fields = {
|
||||||
"instruction": examples["instruction"][i],
|
"instruction": examples["instruction"][i],
|
||||||
"input": examples.get("input", [""] * len(examples["instruction"]))[i],
|
"input": examples.get("input", [""] * len(examples["instruction"]))[
|
||||||
"output": examples["output"][i]
|
i
|
||||||
|
],
|
||||||
|
"output": examples["output"][i],
|
||||||
}
|
}
|
||||||
|
|
||||||
try:
|
try:
|
||||||
text = DEFAULT_ALPACA_TEMPLATE.format(fields["instruction"], fields["input"], fields["output"])
|
text = DEFAULT_ALPACA_TEMPLATE.format(
|
||||||
|
fields["instruction"], fields["input"], fields["output"]
|
||||||
|
)
|
||||||
text += eos_token
|
text += eos_token
|
||||||
texts.append(text)
|
texts.append(text)
|
||||||
except KeyError as e:
|
except KeyError as e:
|
||||||
|
|
@ -284,24 +304,26 @@ def apply_chat_template_to_dataset(
|
||||||
|
|
||||||
try:
|
try:
|
||||||
dataset_map_kwargs = {
|
dataset_map_kwargs = {
|
||||||
'batched': True,
|
"batched": True,
|
||||||
'batch_size': batch_size,
|
"batch_size": batch_size,
|
||||||
}
|
}
|
||||||
|
|
||||||
try:
|
try:
|
||||||
from torch.utils.data import IterableDataset
|
from torch.utils.data import IterableDataset
|
||||||
|
|
||||||
_is_torch_iterable = isinstance(dataset, IterableDataset)
|
_is_torch_iterable = isinstance(dataset, IterableDataset)
|
||||||
except ImportError:
|
except ImportError:
|
||||||
_is_torch_iterable = False
|
_is_torch_iterable = False
|
||||||
|
|
||||||
if not _is_torch_iterable:
|
if not _is_torch_iterable:
|
||||||
from utils.hardware import dataset_map_num_proc
|
from utils.hardware import dataset_map_num_proc
|
||||||
|
|
||||||
if num_proc is None or type(num_proc) is not int:
|
if num_proc is None or type(num_proc) is not int:
|
||||||
num_proc = dataset_map_num_proc()
|
num_proc = dataset_map_num_proc()
|
||||||
else:
|
else:
|
||||||
num_proc = dataset_map_num_proc(num_proc)
|
num_proc = dataset_map_num_proc(num_proc)
|
||||||
dataset_map_kwargs['num_proc'] = num_proc
|
dataset_map_kwargs["num_proc"] = num_proc
|
||||||
dataset_map_kwargs['desc'] = "Applying template to Alpaca format"
|
dataset_map_kwargs["desc"] = "Applying template to Alpaca format"
|
||||||
|
|
||||||
formatted_dataset = dataset.map(formatted_fn, **dataset_map_kwargs)
|
formatted_dataset = dataset.map(formatted_fn, **dataset_map_kwargs)
|
||||||
|
|
||||||
|
|
@ -309,7 +331,7 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": formatted_dataset,
|
"dataset": formatted_dataset,
|
||||||
"success": True,
|
"success": True,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
errors.append(f"Failed to format Alpaca dataset: {e}")
|
errors.append(f"Failed to format Alpaca dataset: {e}")
|
||||||
|
|
@ -317,12 +339,11 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
|
|
||||||
# CHATML FORMATS
|
# CHATML FORMATS
|
||||||
elif final_format in ["chatml_messages", "chatml_conversations"]:
|
elif final_format in ["chatml_messages", "chatml_conversations"]:
|
||||||
|
|
||||||
if not is_standardized:
|
if not is_standardized:
|
||||||
warnings.append("Dataset may not be fully standardized")
|
warnings.append("Dataset may not be fully standardized")
|
||||||
|
|
||||||
|
|
@ -337,13 +358,11 @@ def apply_chat_template_to_dataset(
|
||||||
for convo in convos:
|
for convo in convos:
|
||||||
try:
|
try:
|
||||||
text = tokenizer.apply_chat_template(
|
text = tokenizer.apply_chat_template(
|
||||||
convo,
|
convo, tokenize = False, add_generation_prompt = False
|
||||||
tokenize = False,
|
|
||||||
add_generation_prompt = False
|
|
||||||
)
|
)
|
||||||
|
|
||||||
if remove_bos_prefix:
|
if remove_bos_prefix:
|
||||||
text = text.removeprefix('<bos>')
|
text = text.removeprefix("<bos>")
|
||||||
text += eos_token
|
text += eos_token
|
||||||
|
|
||||||
texts.append(text)
|
texts.append(text)
|
||||||
|
|
@ -357,23 +376,25 @@ def apply_chat_template_to_dataset(
|
||||||
try:
|
try:
|
||||||
try:
|
try:
|
||||||
from torch.utils.data import IterableDataset
|
from torch.utils.data import IterableDataset
|
||||||
|
|
||||||
_is_torch_iterable = isinstance(dataset, IterableDataset)
|
_is_torch_iterable = isinstance(dataset, IterableDataset)
|
||||||
except ImportError:
|
except ImportError:
|
||||||
_is_torch_iterable = False
|
_is_torch_iterable = False
|
||||||
|
|
||||||
dataset_map_kwargs = {
|
dataset_map_kwargs = {
|
||||||
'batched': True,
|
"batched": True,
|
||||||
'batch_size': batch_size,
|
"batch_size": batch_size,
|
||||||
}
|
}
|
||||||
|
|
||||||
if not _is_torch_iterable:
|
if not _is_torch_iterable:
|
||||||
from utils.hardware import dataset_map_num_proc
|
from utils.hardware import dataset_map_num_proc
|
||||||
|
|
||||||
if num_proc is None or type(num_proc) is not int:
|
if num_proc is None or type(num_proc) is not int:
|
||||||
num_proc = dataset_map_num_proc()
|
num_proc = dataset_map_num_proc()
|
||||||
else:
|
else:
|
||||||
num_proc = dataset_map_num_proc(num_proc)
|
num_proc = dataset_map_num_proc(num_proc)
|
||||||
dataset_map_kwargs['num_proc'] = num_proc
|
dataset_map_kwargs["num_proc"] = num_proc
|
||||||
dataset_map_kwargs['desc'] = f"Applying chat template to {final_format}"
|
dataset_map_kwargs["desc"] = f"Applying chat template to {final_format}"
|
||||||
|
|
||||||
# Monitor tqdm progress from dataset.map() and relay to callback
|
# Monitor tqdm progress from dataset.map() and relay to callback
|
||||||
_tqdm_monitor_stop = None
|
_tqdm_monitor_stop = None
|
||||||
|
|
@ -411,7 +432,7 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": formatted_dataset,
|
"dataset": formatted_dataset,
|
||||||
"success": True,
|
"success": True,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
errors.append(f"Failed to format ChatML dataset: {e}")
|
errors.append(f"Failed to format ChatML dataset: {e}")
|
||||||
|
|
@ -419,7 +440,7 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
|
|
||||||
# UNKNOWN FORMAT
|
# UNKNOWN FORMAT
|
||||||
|
|
@ -432,5 +453,5 @@ def apply_chat_template_to_dataset(
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"success": False,
|
"success": False,
|
||||||
"warnings": warnings,
|
"warnings": warnings,
|
||||||
"errors": errors
|
"errors": errors,
|
||||||
}
|
}
|
||||||
|
|
|
||||||
|
|
@ -9,7 +9,7 @@ particularly for VLM/OCR processing.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
from dataclasses import dataclass
|
from dataclasses import dataclass
|
||||||
from typing import Any, List, Optional, Union
|
from typing import Any, List
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
@ -75,7 +75,6 @@ class DeepSeekOCRDataCollator:
|
||||||
Returns:
|
Returns:
|
||||||
dict with input_ids, attention_mask, labels, pixel_values, etc.
|
dict with input_ids, attention_mask, labels, pixel_values, etc.
|
||||||
"""
|
"""
|
||||||
from PIL import Image
|
|
||||||
|
|
||||||
# Extract messages and images
|
# Extract messages and images
|
||||||
all_messages = []
|
all_messages = []
|
||||||
|
|
|
||||||
|
|
@ -38,12 +38,7 @@ from .format_conversion import (
|
||||||
from .chat_templates import (
|
from .chat_templates import (
|
||||||
apply_chat_template_to_dataset,
|
apply_chat_template_to_dataset,
|
||||||
get_dataset_info_summary,
|
get_dataset_info_summary,
|
||||||
get_tokenizer_chat_template,
|
|
||||||
DEFAULT_ALPACA_TEMPLATE,
|
|
||||||
)
|
)
|
||||||
from .vlm_processing import generate_smart_vlm_instruction
|
|
||||||
from .data_collators import DeepSeekOCRDataCollator, VLMDataCollator
|
|
||||||
from .model_mappings import TEMPLATE_TO_MODEL_MAPPER
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
@ -701,7 +696,7 @@ def format_dataset(
|
||||||
}
|
}
|
||||||
|
|
||||||
else:
|
else:
|
||||||
warnings.append(f"Cannot convert unknown format to Alpaca")
|
warnings.append("Cannot convert unknown format to Alpaca")
|
||||||
return {
|
return {
|
||||||
"dataset": dataset,
|
"dataset": dataset,
|
||||||
"detected_format": "unknown",
|
"detected_format": "unknown",
|
||||||
|
|
@ -766,7 +761,7 @@ def format_dataset(
|
||||||
}
|
}
|
||||||
|
|
||||||
else:
|
else:
|
||||||
warnings.append(f"Unknown format, attempting standardization")
|
warnings.append("Unknown format, attempting standardization")
|
||||||
if detected["chat_column"]:
|
if detected["chat_column"]:
|
||||||
try:
|
try:
|
||||||
standardized = standardize_chat_format(
|
standardized = standardize_chat_format(
|
||||||
|
|
@ -914,7 +909,7 @@ def format_and_template_dataset(
|
||||||
f"falling back to auto-detection"
|
f"falling back to auto-detection"
|
||||||
)
|
)
|
||||||
logger.info(
|
logger.info(
|
||||||
f"⚠️ User VLM mapping failed, falling back to auto-detection..."
|
"⚠️ User VLM mapping failed, falling back to auto-detection..."
|
||||||
)
|
)
|
||||||
custom_format_mapping = None # clear so auto-detection runs below
|
custom_format_mapping = None # clear so auto-detection runs below
|
||||||
else:
|
else:
|
||||||
|
|
|
||||||
|
|
@ -41,7 +41,6 @@ def standardize_chat_format(
|
||||||
"""
|
"""
|
||||||
import collections
|
import collections
|
||||||
import itertools
|
import itertools
|
||||||
from datasets import IterableDataset
|
|
||||||
|
|
||||||
# Check if vision tokenizer is used
|
# Check if vision tokenizer is used
|
||||||
is_vlm = False
|
is_vlm = False
|
||||||
|
|
|
||||||
|
|
@ -229,7 +229,7 @@ def detect_custom_format_heuristic(dataset):
|
||||||
if prefix in ["generation", "pass", "inference"]:
|
if prefix in ["generation", "pass", "inference"]:
|
||||||
return True
|
return True
|
||||||
|
|
||||||
if len(col_lower) <= 2 and not col_lower in ["qa", "q", "a"]:
|
if len(col_lower) <= 2 and col_lower not in ["qa", "q", "a"]:
|
||||||
return True
|
return True
|
||||||
|
|
||||||
return False
|
return False
|
||||||
|
|
|
||||||
|
|
@ -19,7 +19,6 @@ import os
|
||||||
import re
|
import re
|
||||||
import textwrap
|
import textwrap
|
||||||
import time
|
import time
|
||||||
from itertools import islice
|
|
||||||
from typing import Any, Optional
|
from typing import Any, Optional
|
||||||
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
@ -105,7 +104,7 @@ def precache_helper_gguf():
|
||||||
finally:
|
finally:
|
||||||
try:
|
try:
|
||||||
enable_progress_bars()
|
enable_progress_bars()
|
||||||
except Exception as e:
|
except Exception:
|
||||||
pass
|
pass
|
||||||
|
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -17,7 +17,6 @@ Usage:
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import platform
|
import platform
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from enum import Enum
|
from enum import Enum
|
||||||
from typing import Optional, Dict, Any
|
from typing import Optional, Dict, Any
|
||||||
|
|
@ -183,7 +182,6 @@ def get_gpu_memory_info() -> Dict[str, Any]:
|
||||||
# ---- MLX path (Apple Silicon) ----
|
# ---- MLX path (Apple Silicon) ----
|
||||||
if device == DeviceType.MLX:
|
if device == DeviceType.MLX:
|
||||||
try:
|
try:
|
||||||
import mlx.core as mx
|
|
||||||
import psutil
|
import psutil
|
||||||
|
|
||||||
# MLX uses unified memory — report system memory as the pool
|
# MLX uses unified memory — report system memory as the pool
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@ from pathlib import Path
|
||||||
from typing import Dict, Any, Optional
|
from typing import Dict, Any, Optional
|
||||||
import json
|
import json
|
||||||
import yaml
|
import yaml
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
from utils.models.model_config import load_model_defaults
|
from utils.models.model_config import load_model_defaults
|
||||||
|
|
|
||||||
|
|
@ -6,7 +6,6 @@ Checkpoint scanning utilities for discovering training runs and their checkpoint
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import json
|
import json
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import List, Optional, Tuple
|
from typing import List, Optional, Tuple
|
||||||
|
|
|
||||||
|
|
@ -18,7 +18,6 @@ from utils.paths import (
|
||||||
resolve_export_dir,
|
resolve_export_dir,
|
||||||
)
|
)
|
||||||
from utils.utils import without_hf_auth
|
from utils.utils import without_hf_auth
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import subprocess
|
import subprocess
|
||||||
|
|
|
||||||
|
|
@ -8,7 +8,6 @@ Path utilities for model and dataset handling
|
||||||
import os
|
import os
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from typing import Optional
|
from typing import Optional
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
|
|
||||||
logger = get_logger(__name__)
|
logger = get_logger(__name__)
|
||||||
|
|
|
||||||
|
|
@ -23,7 +23,6 @@ Strategy:
|
||||||
|
|
||||||
import importlib
|
import importlib
|
||||||
import json
|
import json
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
import os
|
import os
|
||||||
import shutil
|
import shutil
|
||||||
|
|
|
||||||
|
|
@ -6,7 +6,6 @@ Shared backend utilities
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import os
|
import os
|
||||||
import structlog
|
|
||||||
from loggers import get_logger
|
from loggers import get_logger
|
||||||
from contextlib import contextmanager
|
from contextlib import contextmanager
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
|
|
|
||||||
|
|
@ -353,7 +353,7 @@ def pip_install(
|
||||||
)
|
)
|
||||||
if result.returncode == 0:
|
if result.returncode == 0:
|
||||||
return
|
return
|
||||||
print(_red(f" uv failed, falling back to pip..."))
|
print(_red(" uv failed, falling back to pip..."))
|
||||||
if result.stdout:
|
if result.stdout:
|
||||||
print(result.stdout.decode(errors = "replace"))
|
print(result.stdout.decode(errors = "replace"))
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -5,7 +5,6 @@ from __future__ import annotations
|
||||||
import re
|
import re
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
|
|
||||||
import pytest
|
|
||||||
|
|
||||||
REPO_ROOT = Path(__file__).resolve().parents[2]
|
REPO_ROOT = Path(__file__).resolve().parents[2]
|
||||||
INSTALL_SH = REPO_ROOT / "install.sh"
|
INSTALL_SH = REPO_ROOT / "install.sh"
|
||||||
|
|
|
||||||
|
|
@ -21,7 +21,6 @@ from __future__ import annotations
|
||||||
|
|
||||||
import os
|
import os
|
||||||
import shutil
|
import shutil
|
||||||
import signal
|
|
||||||
import subprocess
|
import subprocess
|
||||||
import sys
|
import sys
|
||||||
import textwrap
|
import textwrap
|
||||||
|
|
|
||||||
|
|
@ -2,13 +2,11 @@
|
||||||
|
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import importlib
|
|
||||||
import os
|
import os
|
||||||
import sys
|
import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from unittest import mock
|
from unittest import mock
|
||||||
|
|
||||||
import pytest
|
|
||||||
|
|
||||||
# Add the studio directory so we can import install_python_stack
|
# Add the studio directory so we can import install_python_stack
|
||||||
STUDIO_DIR = Path(__file__).resolve().parents[2] / "studio"
|
STUDIO_DIR = Path(__file__).resolve().parents[2] / "studio"
|
||||||
|
|
|
||||||
|
|
@ -10,7 +10,6 @@ Covers:
|
||||||
|
|
||||||
from __future__ import annotations
|
from __future__ import annotations
|
||||||
|
|
||||||
import importlib
|
|
||||||
import os
|
import os
|
||||||
import re
|
import re
|
||||||
import subprocess
|
import subprocess
|
||||||
|
|
|
||||||
|
|
@ -15,7 +15,6 @@ import ast
|
||||||
import os
|
import os
|
||||||
import shutil
|
import shutil
|
||||||
import subprocess
|
import subprocess
|
||||||
import sys
|
|
||||||
import tempfile
|
import tempfile
|
||||||
import textwrap
|
import textwrap
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
|
|
|
||||||
|
|
@ -1,6 +1,6 @@
|
||||||
from unsloth import FastLanguageModel
|
from unsloth import FastLanguageModel
|
||||||
from unsloth.chat_templates import get_chat_template
|
from unsloth.chat_templates import get_chat_template
|
||||||
from trl import SFTTrainer, SFTConfig
|
from trl import SFTTrainer
|
||||||
from transformers import DataCollatorForSeq2Seq, TrainingArguments
|
from transformers import DataCollatorForSeq2Seq, TrainingArguments
|
||||||
from datasets import load_dataset
|
from datasets import load_dataset
|
||||||
import torch
|
import torch
|
||||||
|
|
|
||||||
|
|
@ -11,7 +11,6 @@ import sys
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
import multiprocessing as mp
|
import multiprocessing as mp
|
||||||
import gc
|
import gc
|
||||||
from multiprocessing import Queue
|
|
||||||
|
|
||||||
REPO_ROOT = Path(__file__).parents[3]
|
REPO_ROOT = Path(__file__).parents[3]
|
||||||
sys.path.insert(0, str(REPO_ROOT))
|
sys.path.insert(0, str(REPO_ROOT))
|
||||||
|
|
|
||||||
|
|
@ -1,6 +1,4 @@
|
||||||
from unsloth import FastLanguageModel
|
from unsloth import FastLanguageModel
|
||||||
from transformers import AutoModelForCausalLM
|
|
||||||
from peft import PeftModel
|
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
import sys
|
import sys
|
||||||
import warnings
|
import warnings
|
||||||
|
|
|
||||||
|
|
@ -1,6 +1,5 @@
|
||||||
from unsloth import FastLanguageModel, FastModel
|
from unsloth import FastModel
|
||||||
from transformers import AutoModelForCausalLM, WhisperForConditionalGeneration
|
from transformers import WhisperForConditionalGeneration
|
||||||
from peft import PeftModel
|
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
import sys
|
import sys
|
||||||
import warnings
|
import warnings
|
||||||
|
|
|
||||||
|
|
@ -28,7 +28,7 @@ train_dataset = dataset.select(range(2000))
|
||||||
# To select the next 200 examples for evaluation
|
# To select the next 200 examples for evaluation
|
||||||
eval_dataset = dataset.select(range(2000, 2200))
|
eval_dataset = dataset.select(range(2000, 2200))
|
||||||
|
|
||||||
print(f"✅ Dataset loaded successfully!")
|
print("✅ Dataset loaded successfully!")
|
||||||
print(f" 📈 Training samples: {len(train_dataset)}")
|
print(f" 📈 Training samples: {len(train_dataset)}")
|
||||||
print(f" 📊 Evaluation samples: {len(eval_dataset)}")
|
print(f" 📊 Evaluation samples: {len(eval_dataset)}")
|
||||||
|
|
||||||
|
|
@ -110,10 +110,10 @@ try:
|
||||||
loftq_config = None, # And LoftQ
|
loftq_config = None, # And LoftQ
|
||||||
)
|
)
|
||||||
print("✅ LoRA configuration applied successfully!")
|
print("✅ LoRA configuration applied successfully!")
|
||||||
print(f" 🎯 LoRA rank (r): 16")
|
print(" 🎯 LoRA rank (r): 16")
|
||||||
print(f" 📊 LoRA alpha: 32")
|
print(" 📊 LoRA alpha: 32")
|
||||||
print(f" 🔍 Vision layers: Enabled")
|
print(" 🔍 Vision layers: Enabled")
|
||||||
print(f" 💬 Language layers: Enabled")
|
print(" 💬 Language layers: Enabled")
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print(f"❌ Failed to apply LoRA configuration: {e}")
|
print(f"❌ Failed to apply LoRA configuration: {e}")
|
||||||
raise
|
raise
|
||||||
|
|
@ -165,10 +165,10 @@ try:
|
||||||
),
|
),
|
||||||
)
|
)
|
||||||
print("✅ Trainer setup completed!")
|
print("✅ Trainer setup completed!")
|
||||||
print(f" 📦 Batch size: 2")
|
print(" 📦 Batch size: 2")
|
||||||
print(f" 🔄 Gradient accumulation steps: 4")
|
print(" 🔄 Gradient accumulation steps: 4")
|
||||||
print(f" 📈 Max training steps: 10")
|
print(" 📈 Max training steps: 10")
|
||||||
print(f" 🎯 Learning rate: 2e-4")
|
print(" 🎯 Learning rate: 2e-4")
|
||||||
print(f" 💾 Precision: {'BF16' if is_bf16_supported() else 'FP16'}")
|
print(f" 💾 Precision: {'BF16' if is_bf16_supported() else 'FP16'}")
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print(f"❌ Failed to setup trainer: {e}")
|
print(f"❌ Failed to setup trainer: {e}")
|
||||||
|
|
|
||||||
|
|
@ -29,7 +29,7 @@ train_dataset = dataset.select(range(2000))
|
||||||
# To select the next 200 examples for evaluation
|
# To select the next 200 examples for evaluation
|
||||||
eval_dataset = dataset.select(range(2000, 2200))
|
eval_dataset = dataset.select(range(2000, 2200))
|
||||||
|
|
||||||
print(f"✅ Dataset loaded successfully!")
|
print("✅ Dataset loaded successfully!")
|
||||||
print(f" 📈 Training samples: {len(train_dataset)}")
|
print(f" 📈 Training samples: {len(train_dataset)}")
|
||||||
print(f" 📊 Evaluation samples: {len(eval_dataset)}")
|
print(f" 📊 Evaluation samples: {len(eval_dataset)}")
|
||||||
|
|
||||||
|
|
@ -111,10 +111,10 @@ try:
|
||||||
loftq_config = None, # And LoftQ
|
loftq_config = None, # And LoftQ
|
||||||
)
|
)
|
||||||
print("✅ LoRA configuration applied successfully!")
|
print("✅ LoRA configuration applied successfully!")
|
||||||
print(f" 🎯 LoRA rank (r): 16")
|
print(" 🎯 LoRA rank (r): 16")
|
||||||
print(f" 📊 LoRA alpha: 32")
|
print(" 📊 LoRA alpha: 32")
|
||||||
print(f" 🔍 Vision layers: Enabled")
|
print(" 🔍 Vision layers: Enabled")
|
||||||
print(f" 💬 Language layers: Enabled")
|
print(" 💬 Language layers: Enabled")
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print(f"❌ Failed to apply LoRA configuration: {e}")
|
print(f"❌ Failed to apply LoRA configuration: {e}")
|
||||||
raise
|
raise
|
||||||
|
|
@ -166,10 +166,10 @@ try:
|
||||||
),
|
),
|
||||||
)
|
)
|
||||||
print("✅ Trainer setup completed!")
|
print("✅ Trainer setup completed!")
|
||||||
print(f" 📦 Batch size: 2")
|
print(" 📦 Batch size: 2")
|
||||||
print(f" 🔄 Gradient accumulation steps: 4")
|
print(" 🔄 Gradient accumulation steps: 4")
|
||||||
print(f" 📈 Max training steps: 10")
|
print(" 📈 Max training steps: 10")
|
||||||
print(f" 🎯 Learning rate: 2e-4")
|
print(" 🎯 Learning rate: 2e-4")
|
||||||
print(f" 💾 Precision: {'BF16' if is_bf16_supported() else 'FP16'}")
|
print(f" 💾 Precision: {'BF16' if is_bf16_supported() else 'FP16'}")
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print(f"❌ Failed to setup trainer: {e}")
|
print(f"❌ Failed to setup trainer: {e}")
|
||||||
|
|
|
||||||
|
|
@ -2,9 +2,6 @@
|
||||||
|
|
||||||
from unsloth import FastVisionModel
|
from unsloth import FastVisionModel
|
||||||
|
|
||||||
import torch
|
|
||||||
from qwen_vl_utils import process_vision_info
|
|
||||||
import os
|
|
||||||
from datasets import load_dataset
|
from datasets import load_dataset
|
||||||
from trl import SFTTrainer, SFTConfig
|
from trl import SFTTrainer, SFTConfig
|
||||||
|
|
||||||
|
|
@ -20,7 +17,6 @@ from tests.utils.ocr_eval import OCRModelEvaluator
|
||||||
|
|
||||||
|
|
||||||
## Dataset Preparation
|
## Dataset Preparation
|
||||||
from datasets import load_dataset
|
|
||||||
|
|
||||||
dataset = load_dataset("lbourdois/OCR-liboaccn-OPUS-MIT-5M-clean", "en", split = "train")
|
dataset = load_dataset("lbourdois/OCR-liboaccn-OPUS-MIT-5M-clean", "en", split = "train")
|
||||||
# To select the first 2000 examples
|
# To select the first 2000 examples
|
||||||
|
|
@ -66,12 +62,6 @@ train_dataset = [format_data(sample) for sample in train_dataset]
|
||||||
eval_dataset = [format_data(sample) for sample in eval_dataset]
|
eval_dataset = [format_data(sample) for sample in eval_dataset]
|
||||||
|
|
||||||
## Setup OCR main evaluation function and helpers
|
## Setup OCR main evaluation function and helpers
|
||||||
import os
|
|
||||||
import torch
|
|
||||||
from tqdm import tqdm
|
|
||||||
import pandas as pd
|
|
||||||
from jiwer import wer, cer
|
|
||||||
from qwen_vl_utils import process_vision_info
|
|
||||||
|
|
||||||
#
|
#
|
||||||
ocr_evaluator = OCRModelEvaluator()
|
ocr_evaluator = OCRModelEvaluator()
|
||||||
|
|
|
||||||
|
|
@ -2,9 +2,6 @@
|
||||||
|
|
||||||
from unsloth import FastVisionModel
|
from unsloth import FastVisionModel
|
||||||
|
|
||||||
import torch
|
|
||||||
from qwen_vl_utils import process_vision_info
|
|
||||||
import os
|
|
||||||
from datasets import load_dataset
|
from datasets import load_dataset
|
||||||
from trl import SFTTrainer, SFTConfig
|
from trl import SFTTrainer, SFTConfig
|
||||||
|
|
||||||
|
|
@ -20,7 +17,6 @@ from tests.utils.ocr_eval import OCRModelEvaluator
|
||||||
|
|
||||||
|
|
||||||
## Dataset Preparation
|
## Dataset Preparation
|
||||||
from datasets import load_dataset
|
|
||||||
|
|
||||||
dataset = load_dataset("lbourdois/OCR-liboaccn-OPUS-MIT-5M-clean", "en", split = "train")
|
dataset = load_dataset("lbourdois/OCR-liboaccn-OPUS-MIT-5M-clean", "en", split = "train")
|
||||||
# To select the first 2000 examples
|
# To select the first 2000 examples
|
||||||
|
|
@ -66,12 +62,6 @@ train_dataset = [format_data(sample) for sample in train_dataset]
|
||||||
eval_dataset = [format_data(sample) for sample in eval_dataset]
|
eval_dataset = [format_data(sample) for sample in eval_dataset]
|
||||||
|
|
||||||
## Setup OCR main evaluation function and helpers
|
## Setup OCR main evaluation function and helpers
|
||||||
import os
|
|
||||||
import torch
|
|
||||||
from tqdm import tqdm
|
|
||||||
import pandas as pd
|
|
||||||
from jiwer import wer, cer
|
|
||||||
from qwen_vl_utils import process_vision_info
|
|
||||||
|
|
||||||
#
|
#
|
||||||
ocr_evaluator = OCRModelEvaluator()
|
ocr_evaluator = OCRModelEvaluator()
|
||||||
|
|
|
||||||
|
|
@ -1,6 +1,5 @@
|
||||||
import importlib.util
|
import importlib.util
|
||||||
import io
|
import io
|
||||||
import json
|
|
||||||
import os
|
import os
|
||||||
import sys
|
import sys
|
||||||
import tarfile
|
import tarfile
|
||||||
|
|
|
||||||
|
|
@ -14,13 +14,11 @@ Run: pytest tests/studio/install/test_pr4562_bugfixes.py -v
|
||||||
"""
|
"""
|
||||||
|
|
||||||
import importlib.util
|
import importlib.util
|
||||||
import json
|
|
||||||
import os
|
import os
|
||||||
import subprocess
|
import subprocess
|
||||||
import sys
|
import sys
|
||||||
import textwrap
|
import textwrap
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
from unittest.mock import patch
|
|
||||||
|
|
||||||
import pytest
|
import pytest
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -7,7 +7,6 @@ Tests basic functionality without heavy dependencies.
|
||||||
import sys
|
import sys
|
||||||
import os
|
import os
|
||||||
import tempfile
|
import tempfile
|
||||||
from pathlib import Path
|
|
||||||
import importlib.util
|
import importlib.util
|
||||||
|
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -207,7 +207,7 @@ def evaluate_model_aime(
|
||||||
|
|
||||||
print(f"\n{'='*70}")
|
print(f"\n{'='*70}")
|
||||||
print(f"🧮 AIME EVALUATION - {model_type.upper()} MODEL")
|
print(f"🧮 AIME EVALUATION - {model_type.upper()} MODEL")
|
||||||
print(f"Combined Dataset: test2024 + test2025-I + test2025-II")
|
print("Combined Dataset: test2024 + test2025-I + test2025-II")
|
||||||
print(f"{'='*70}")
|
print(f"{'='*70}")
|
||||||
|
|
||||||
# Load combined AIME dataset
|
# Load combined AIME dataset
|
||||||
|
|
@ -244,7 +244,7 @@ def evaluate_model_aime(
|
||||||
seed = seed,
|
seed = seed,
|
||||||
)
|
)
|
||||||
|
|
||||||
print(f"\n🔧 Configuration:")
|
print("\n🔧 Configuration:")
|
||||||
print(f" Temperature: {temperature}")
|
print(f" Temperature: {temperature}")
|
||||||
print(f" Samples per question: {n_sampling}")
|
print(f" Samples per question: {n_sampling}")
|
||||||
print(f" Max tokens: {max_tokens}")
|
print(f" Max tokens: {max_tokens}")
|
||||||
|
|
@ -421,28 +421,28 @@ def evaluate_model_aime(
|
||||||
print(f"📊 AIME EVALUATION RESULTS - {model_type.upper()}")
|
print(f"📊 AIME EVALUATION RESULTS - {model_type.upper()}")
|
||||||
print(f"{'='*70}")
|
print(f"{'='*70}")
|
||||||
|
|
||||||
print(f"\n🎯 Overall Performance:")
|
print("\n🎯 Overall Performance:")
|
||||||
print(f" Total problems: {total_problems:>6}")
|
print(f" Total problems: {total_problems:>6}")
|
||||||
print(
|
print(
|
||||||
f" Correct answers: {correct_answers:>6}/{total_problems} ({accuracy:>5.1f}%)"
|
f" Correct answers: {correct_answers:>6}/{total_problems} ({accuracy:>5.1f}%)"
|
||||||
)
|
)
|
||||||
print(f" Pass@{n_sampling}: {pass_at_k:>10.1f}%")
|
print(f" Pass@{n_sampling}: {pass_at_k:>10.1f}%")
|
||||||
|
|
||||||
print(f"\n📈 Performance by Dataset:")
|
print("\n📈 Performance by Dataset:")
|
||||||
for source, stats in source_stats.items():
|
for source, stats in source_stats.items():
|
||||||
source_acc = source_accuracies[source]
|
source_acc = source_accuracies[source]
|
||||||
print(
|
print(
|
||||||
f" {source:>12}: {stats['correct']:>3}/{stats['total']:>3} ({source_acc:>5.1f}%)"
|
f" {source:>12}: {stats['correct']:>3}/{stats['total']:>3} ({source_acc:>5.1f}%)"
|
||||||
)
|
)
|
||||||
|
|
||||||
print(f"\n🔧 Configuration:")
|
print("\n🔧 Configuration:")
|
||||||
print(f" Temperature: {temperature}")
|
print(f" Temperature: {temperature}")
|
||||||
print(f" Samples per problem: {n_sampling}")
|
print(f" Samples per problem: {n_sampling}")
|
||||||
print(f" Max tokens: {max_tokens}")
|
print(f" Max tokens: {max_tokens}")
|
||||||
print(f" Top-p: {top_p}")
|
print(f" Top-p: {top_p}")
|
||||||
print(f" Seed: {seed}")
|
print(f" Seed: {seed}")
|
||||||
|
|
||||||
print(f"\n📝 Token Statistics:")
|
print("\n📝 Token Statistics:")
|
||||||
print(f" Avg input tokens: {results['avg_input_tokens']:>10.1f}")
|
print(f" Avg input tokens: {results['avg_input_tokens']:>10.1f}")
|
||||||
print(f" Avg output tokens: {results['avg_output_tokens']:>10.1f}")
|
print(f" Avg output tokens: {results['avg_output_tokens']:>10.1f}")
|
||||||
print(f" Max input tokens: {results['max_input_tokens']:>10}")
|
print(f" Max input tokens: {results['max_input_tokens']:>10}")
|
||||||
|
|
|
||||||
|
|
@ -12,9 +12,8 @@
|
||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
import os
|
from contextlib import nullcontext
|
||||||
from contextlib import contextmanager, nullcontext
|
from typing import Callable
|
||||||
from typing import Callable, Optional
|
|
||||||
|
|
||||||
import bitsandbytes as bnb
|
import bitsandbytes as bnb
|
||||||
import torch
|
import torch
|
||||||
|
|
|
||||||
|
|
@ -348,7 +348,7 @@ class OCRModelEvaluator:
|
||||||
plt.savefig("ocr_model_comparison.png")
|
plt.savefig("ocr_model_comparison.png")
|
||||||
plt.show()
|
plt.show()
|
||||||
|
|
||||||
print(f"\nVisualization saved to ocr_model_comparison.png")
|
print("\nVisualization saved to ocr_model_comparison.png")
|
||||||
|
|
||||||
def get_comparison_results(self) -> Dict[str, Dict[str, float]]:
|
def get_comparison_results(self) -> Dict[str, Dict[str, float]]:
|
||||||
"""Get the current comparison results."""
|
"""Get the current comparison results."""
|
||||||
|
|
|
||||||
|
|
@ -98,10 +98,10 @@ def require_package(package_name, executable_name = None):
|
||||||
for pm_name, cmd in install_commands.items():
|
for pm_name, cmd in install_commands.items():
|
||||||
print(f" {pm_name}: {cmd}")
|
print(f" {pm_name}: {cmd}")
|
||||||
|
|
||||||
print(f"\nAlternatively, install with conda:")
|
print("\nAlternatively, install with conda:")
|
||||||
print(f" conda install -c conda-forge {package_name}")
|
print(f" conda install -c conda-forge {package_name}")
|
||||||
|
|
||||||
print(f"\nPlease install the required package and run the script again.")
|
print("\nPlease install the required package and run the script again.")
|
||||||
sys.exit(1)
|
sys.exit(1)
|
||||||
|
|
||||||
|
|
||||||
|
|
@ -120,9 +120,9 @@ def require_python_package(package_name, import_name = None, pip_name = None):
|
||||||
print(f"❌ Error: Python package '{package_name}' is not installed")
|
print(f"❌ Error: Python package '{package_name}' is not installed")
|
||||||
print(f"\nPlease install {package_name} using pip:")
|
print(f"\nPlease install {package_name} using pip:")
|
||||||
print(f" pip install {pip_name}")
|
print(f" pip install {pip_name}")
|
||||||
print(f" # or with conda:")
|
print(" # or with conda:")
|
||||||
print(f" conda install {pip_name}")
|
print(f" conda install {pip_name}")
|
||||||
print(f"\nAfter installation, run this script again.")
|
print("\nAfter installation, run this script again.")
|
||||||
sys.exit(1)
|
sys.exit(1)
|
||||||
else:
|
else:
|
||||||
print(f"✓ Python package '{package_name}' is installed")
|
print(f"✓ Python package '{package_name}' is installed")
|
||||||
|
|
|
||||||
|
|
@ -14,7 +14,6 @@
|
||||||
#
|
#
|
||||||
# Tests for Q-GaLore integration (unsloth/optimizers/).
|
# Tests for Q-GaLore integration (unsloth/optimizers/).
|
||||||
|
|
||||||
import pytest
|
|
||||||
import sys
|
import sys
|
||||||
import os
|
import os
|
||||||
import torch
|
import torch
|
||||||
|
|
|
||||||
|
|
@ -101,8 +101,6 @@ def run(args):
|
||||||
return {"text": texts}
|
return {"text": texts}
|
||||||
|
|
||||||
def load_dataset_smart(args):
|
def load_dataset_smart(args):
|
||||||
from transformers.utils import strtobool
|
|
||||||
|
|
||||||
if args.raw_text_file:
|
if args.raw_text_file:
|
||||||
# Use raw text loader
|
# Use raw text loader
|
||||||
loader = RawTextDataLoader(tokenizer, args.chunk_size, args.stride)
|
loader = RawTextDataLoader(tokenizer, args.chunk_size, args.stride)
|
||||||
|
|
|
||||||
|
|
@ -12,9 +12,15 @@
|
||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
import warnings, importlib, sys
|
import warnings
|
||||||
|
import importlib
|
||||||
|
import sys
|
||||||
from packaging.version import Version
|
from packaging.version import Version
|
||||||
import os, re, subprocess, inspect, functools
|
import os
|
||||||
|
import re
|
||||||
|
import subprocess
|
||||||
|
import inspect
|
||||||
|
import functools
|
||||||
import numpy as np
|
import numpy as np
|
||||||
|
|
||||||
# Log Unsloth is being used
|
# Log Unsloth is being used
|
||||||
|
|
@ -105,7 +111,7 @@ try:
|
||||||
import unsloth_zoo
|
import unsloth_zoo
|
||||||
except PackageNotFoundError:
|
except PackageNotFoundError:
|
||||||
raise ImportError(
|
raise ImportError(
|
||||||
f"Unsloth: Please install unsloth_zoo via `pip install unsloth_zoo` then retry!"
|
"Unsloth: Please install unsloth_zoo via `pip install unsloth_zoo` then retry!"
|
||||||
)
|
)
|
||||||
except:
|
except:
|
||||||
raise
|
raise
|
||||||
|
|
|
||||||
|
|
@ -12,32 +12,60 @@
|
||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
try: import torch
|
try:
|
||||||
except: raise ImportError('Install torch via `pip install torch`')
|
import torch
|
||||||
|
except:
|
||||||
|
raise ImportError("Install torch via `pip install torch`")
|
||||||
from packaging.version import Version as V
|
from packaging.version import Version as V
|
||||||
import re
|
import re
|
||||||
|
|
||||||
v = V(re.match(r"[0-9\.]{3,}", torch.__version__).group(0))
|
v = V(re.match(r"[0-9\.]{3,}", torch.__version__).group(0))
|
||||||
cuda = str(torch.version.cuda)
|
cuda = str(torch.version.cuda)
|
||||||
is_ampere = torch.cuda.get_device_capability()[0] >= 8
|
is_ampere = torch.cuda.get_device_capability()[0] >= 8
|
||||||
USE_ABI = torch._C._GLIBCXX_USE_CXX11_ABI
|
USE_ABI = torch._C._GLIBCXX_USE_CXX11_ABI
|
||||||
if cuda not in ("11.8", "12.1", "12.4", "12.6", "12.8", "13.0"): raise RuntimeError(f"CUDA = {cuda} not supported!")
|
if cuda not in ("11.8", "12.1", "12.4", "12.6", "12.8", "13.0"):
|
||||||
if v <= V('2.1.0'): raise RuntimeError(f"Torch = {v} too old!")
|
raise RuntimeError(f"CUDA = {cuda} not supported!")
|
||||||
elif v <= V('2.1.1'): x = 'cu{}{}-torch211'
|
if v <= V("2.1.0"):
|
||||||
elif v <= V('2.1.2'): x = 'cu{}{}-torch212'
|
raise RuntimeError(f"Torch = {v} too old!")
|
||||||
elif v < V('2.3.0'): x = 'cu{}{}-torch220'
|
elif v <= V("2.1.1"):
|
||||||
elif v < V('2.4.0'): x = 'cu{}{}-torch230'
|
x = "cu{}{}-torch211"
|
||||||
elif v < V('2.5.0'): x = 'cu{}{}-torch240'
|
elif v <= V("2.1.2"):
|
||||||
elif v < V('2.5.1'): x = 'cu{}{}-torch250'
|
x = "cu{}{}-torch212"
|
||||||
elif v <= V('2.5.1'): x = 'cu{}{}-torch251'
|
elif v < V("2.3.0"):
|
||||||
elif v < V('2.7.0'): x = 'cu{}{}-torch260'
|
x = "cu{}{}-torch220"
|
||||||
elif v < V('2.7.9'): x = 'cu{}{}-torch270'
|
elif v < V("2.4.0"):
|
||||||
elif v < V('2.8.0'): x = 'cu{}{}-torch271'
|
x = "cu{}{}-torch230"
|
||||||
elif v < V('2.8.9'): x = 'cu{}{}-torch280'
|
elif v < V("2.5.0"):
|
||||||
elif v < V('2.9.1'): x = 'cu{}{}-torch290'
|
x = "cu{}{}-torch240"
|
||||||
elif v < V('2.9.2'): x = 'cu{}{}-torch291'
|
elif v < V("2.5.1"):
|
||||||
elif v < V('2.10.1'): x = 'cu{}{}-torch2100'
|
x = "cu{}{}-torch250"
|
||||||
else: raise RuntimeError(f"Torch = {v} too new!")
|
elif v <= V("2.5.1"):
|
||||||
if v > V('2.6.9') and cuda not in ("11.8", "12.6", "12.8", "13.0"): raise RuntimeError(f"CUDA = {cuda} not supported!")
|
x = "cu{}{}-torch251"
|
||||||
if v >= V('2.10.0') and cuda not in ("12.6", "12.8", "13.0"): raise RuntimeError(f"Torch 2.10 requires CUDA 12.6, 12.8, or 13.0! Got CUDA = {cuda}")
|
elif v < V("2.7.0"):
|
||||||
x = x.format(cuda.replace(".", ""), "-ampere" if False else "") # is_ampere is broken due to flash-attn
|
x = "cu{}{}-torch260"
|
||||||
print(f'pip install --upgrade pip && pip install --no-deps git+https://github.com/unslothai/unsloth-zoo.git && pip install "unsloth[{x}] @ git+https://github.com/unslothai/unsloth.git" --no-build-isolation')
|
elif v < V("2.7.9"):
|
||||||
|
x = "cu{}{}-torch270"
|
||||||
|
elif v < V("2.8.0"):
|
||||||
|
x = "cu{}{}-torch271"
|
||||||
|
elif v < V("2.8.9"):
|
||||||
|
x = "cu{}{}-torch280"
|
||||||
|
elif v < V("2.9.1"):
|
||||||
|
x = "cu{}{}-torch290"
|
||||||
|
elif v < V("2.9.2"):
|
||||||
|
x = "cu{}{}-torch291"
|
||||||
|
elif v < V("2.10.1"):
|
||||||
|
x = "cu{}{}-torch2100"
|
||||||
|
else:
|
||||||
|
raise RuntimeError(f"Torch = {v} too new!")
|
||||||
|
if v > V("2.6.9") and cuda not in ("11.8", "12.6", "12.8", "13.0"):
|
||||||
|
raise RuntimeError(f"CUDA = {cuda} not supported!")
|
||||||
|
if v >= V("2.10.0") and cuda not in ("12.6", "12.8", "13.0"):
|
||||||
|
raise RuntimeError(
|
||||||
|
f"Torch 2.10 requires CUDA 12.6, 12.8, or 13.0! Got CUDA = {cuda}"
|
||||||
|
)
|
||||||
|
x = x.format(
|
||||||
|
cuda.replace(".", ""), "-ampere" if False else ""
|
||||||
|
) # is_ampere is broken due to flash-attn
|
||||||
|
print(
|
||||||
|
f'pip install --upgrade pip && pip install --no-deps git+https://github.com/unslothai/unsloth-zoo.git && pip install "unsloth[{x}] @ git+https://github.com/unslothai/unsloth.git" --no-build-isolation'
|
||||||
|
)
|
||||||
|
|
|
||||||
File diff suppressed because it is too large
Load diff
|
|
@ -12,11 +12,9 @@
|
||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
import os
|
|
||||||
import re
|
import re
|
||||||
import json
|
import json
|
||||||
import csv
|
import csv
|
||||||
from typing import List, Dict, Any, Union, Optional
|
|
||||||
from datasets import Dataset
|
from datasets import Dataset
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -25,7 +25,6 @@ os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
|
||||||
import requests
|
import requests
|
||||||
import torch
|
import torch
|
||||||
import gc
|
import gc
|
||||||
import time
|
|
||||||
import re
|
import re
|
||||||
from unsloth_zoo.log import logger
|
from unsloth_zoo.log import logger
|
||||||
import numpy as np
|
import numpy as np
|
||||||
|
|
|
||||||
|
|
@ -75,7 +75,7 @@ class HideLoggingMessage(logging.Filter):
|
||||||
self.text = text
|
self.text = text
|
||||||
|
|
||||||
def filter(self, x):
|
def filter(self, x):
|
||||||
return not (self.text in x.getMessage())
|
return self.text not in x.getMessage()
|
||||||
|
|
||||||
|
|
||||||
class HidePrintMessage:
|
class HidePrintMessage:
|
||||||
|
|
@ -1310,7 +1310,7 @@ def disable_broken_wandb():
|
||||||
return # wandb not installed, nothing to do
|
return # wandb not installed, nothing to do
|
||||||
|
|
||||||
try:
|
try:
|
||||||
import wandb
|
pass
|
||||||
except Exception:
|
except Exception:
|
||||||
# wandb is installed but broken - patch all checkers to skip it
|
# wandb is installed but broken - patch all checkers to skip it
|
||||||
logger.info(
|
logger.info(
|
||||||
|
|
|
||||||
|
|
@ -23,12 +23,10 @@ from .utils import (
|
||||||
torch_gpu_device,
|
torch_gpu_device,
|
||||||
is_cdna,
|
is_cdna,
|
||||||
)
|
)
|
||||||
from transformers.models.llama.modeling_llama import logger
|
|
||||||
from unsloth_zoo.utils import Version
|
from unsloth_zoo.utils import Version
|
||||||
|
|
||||||
from unsloth_zoo.loss_utils import (
|
from unsloth_zoo.loss_utils import (
|
||||||
patch_loss_functions as _patch_loss_functions,
|
patch_loss_functions as _patch_loss_functions,
|
||||||
post_patch_loss_function,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -18,7 +18,6 @@ from .utils import (
|
||||||
fast_dequantize,
|
fast_dequantize,
|
||||||
QUANT_STATE,
|
QUANT_STATE,
|
||||||
get_lora_parameters,
|
get_lora_parameters,
|
||||||
get_lora_parameters_bias,
|
|
||||||
matmul_lora,
|
matmul_lora,
|
||||||
torch_amp_custom_fwd,
|
torch_amp_custom_fwd,
|
||||||
torch_amp_custom_bwd,
|
torch_amp_custom_bwd,
|
||||||
|
|
|
||||||
|
|
@ -13,8 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
import torch
|
import torch
|
||||||
from functools import lru_cache
|
|
||||||
from transformers.models.llama.modeling_llama import logger
|
|
||||||
import os
|
import os
|
||||||
|
|
||||||
torch_compile_options = {
|
torch_compile_options = {
|
||||||
|
|
@ -80,7 +78,8 @@ else:
|
||||||
# See https://github.com/pytorch-labs/attention-gym/blob/main/examples/flex_attn.ipynb
|
# See https://github.com/pytorch-labs/attention-gym/blob/main/examples/flex_attn.ipynb
|
||||||
# for more examples
|
# for more examples
|
||||||
# BSD 3-Clause License Copyright (c) 2023, Driss Guessous, Horace He et al
|
# BSD 3-Clause License Copyright (c) 2023, Driss Guessous, Horace He et al
|
||||||
import functools, math
|
import functools
|
||||||
|
import math
|
||||||
|
|
||||||
def generate_tanh_softcap(t):
|
def generate_tanh_softcap(t):
|
||||||
def tanh_softcap(x, b, h, q_idx, kv_idx):
|
def tanh_softcap(x, b, h, q_idx, kv_idx):
|
||||||
|
|
|
||||||
|
|
@ -13,11 +13,8 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
import os
|
import os
|
||||||
import torch
|
import torch
|
||||||
import torch.nn as nn
|
|
||||||
import triton
|
import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
from torch.nn import functional as F
|
|
||||||
import math
|
|
||||||
from unsloth_zoo.utils import Version
|
from unsloth_zoo.utils import Version
|
||||||
from unsloth_zoo.log import logger
|
from unsloth_zoo.log import logger
|
||||||
from unsloth_zoo.temporary_patches.common import torch_compile
|
from unsloth_zoo.temporary_patches.common import torch_compile
|
||||||
|
|
@ -588,7 +585,7 @@ try:
|
||||||
_has_fbgemm = test_has_fbgemm()
|
_has_fbgemm = test_has_fbgemm()
|
||||||
if _has_fbgemm:
|
if _has_fbgemm:
|
||||||
os.environ["UNSLOTH_HAS_FBGEMM"] = "1"
|
os.environ["UNSLOTH_HAS_FBGEMM"] = "1"
|
||||||
logger.info(f"Using fbgemm_gpu block quantized FP8 matmul")
|
logger.info("Using fbgemm_gpu block quantized FP8 matmul")
|
||||||
fp8_block_quant_linear = fp8_fbgemm_block_linear
|
fp8_block_quant_linear = fp8_fbgemm_block_linear
|
||||||
else:
|
else:
|
||||||
os.environ["UNSLOTH_HAS_FBGEMM"] = "0"
|
os.environ["UNSLOTH_HAS_FBGEMM"] = "0"
|
||||||
|
|
|
||||||
|
|
@ -16,7 +16,6 @@ import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
import torch
|
import torch
|
||||||
from .utils import (
|
from .utils import (
|
||||||
calculate_settings,
|
|
||||||
triton_tanh,
|
triton_tanh,
|
||||||
torch_gpu_device,
|
torch_gpu_device,
|
||||||
)
|
)
|
||||||
|
|
|
||||||
|
|
@ -17,9 +17,6 @@ import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
import torch
|
import torch
|
||||||
from .utils import calculate_settings, torch_gpu_device
|
from .utils import calculate_settings, torch_gpu_device
|
||||||
from unsloth_zoo.patching_utils import (
|
|
||||||
patch_layernorm,
|
|
||||||
)
|
|
||||||
|
|
||||||
|
|
||||||
@triton.jit
|
@triton.jit
|
||||||
|
|
|
||||||
|
|
@ -23,9 +23,8 @@ import json
|
||||||
import logging
|
import logging
|
||||||
import os
|
import os
|
||||||
import time
|
import time
|
||||||
from typing import Dict, List, Optional, Tuple, Any
|
from typing import Dict, Optional, Tuple, Any
|
||||||
import torch
|
import torch
|
||||||
import triton
|
|
||||||
|
|
||||||
logger = logging.getLogger(__name__)
|
logger = logging.getLogger(__name__)
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -178,7 +178,7 @@ def save_autotune_results(autotune_cache, mode, ref_time, fused_time, results_di
|
||||||
|
|
||||||
for key, config in autotune_cache.items():
|
for key, config in autotune_cache.items():
|
||||||
key = [
|
key = [
|
||||||
str(k) if not "torch" in str(k) else str(k.split("torch.")[-1]) for k in key
|
str(k) if "torch" not in str(k) else str(k.split("torch.")[-1]) for k in key
|
||||||
]
|
]
|
||||||
filename = "_".join(key)
|
filename = "_".join(key)
|
||||||
save_path = f"{save_dir}/{filename}.json"
|
save_path = f"{save_dir}/{filename}.json"
|
||||||
|
|
|
||||||
|
|
@ -3,7 +3,6 @@
|
||||||
|
|
||||||
import logging
|
import logging
|
||||||
import warnings
|
import warnings
|
||||||
from dataclasses import asdict
|
|
||||||
from unsloth import DEVICE_TYPE
|
from unsloth import DEVICE_TYPE
|
||||||
|
|
||||||
import torch
|
import torch
|
||||||
|
|
|
||||||
|
|
@ -336,7 +336,6 @@ def exceeds_smem_capacity(
|
||||||
|
|
||||||
|
|
||||||
def common_prune_criteria(config: triton.Config, kwargs: dict, dtype):
|
def common_prune_criteria(config: triton.Config, kwargs: dict, dtype):
|
||||||
from ..interface import supports_tma
|
|
||||||
from .tuning import get_device_properties
|
from .tuning import get_device_properties
|
||||||
|
|
||||||
smem_size = get_device_properties().SIZE_SMEM
|
smem_size = get_device_properties().SIZE_SMEM
|
||||||
|
|
|
||||||
|
|
@ -1,7 +1,6 @@
|
||||||
# SPDX-License-Identifier: GNU Affero General Public License v3.0
|
# SPDX-License-Identifier: GNU Affero General Public License v3.0
|
||||||
# Copyright 2023-present the Unsloth team. All rights reserved.
|
# Copyright 2023-present the Unsloth team. All rights reserved.
|
||||||
|
|
||||||
import torch
|
|
||||||
import triton
|
import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -1,7 +1,6 @@
|
||||||
# SPDX-License-Identifier: GNU Affero General Public License v3.0
|
# SPDX-License-Identifier: GNU Affero General Public License v3.0
|
||||||
# Copyright 2023-present the Unsloth team. All rights reserved.
|
# Copyright 2023-present the Unsloth team. All rights reserved.
|
||||||
|
|
||||||
import torch
|
|
||||||
import triton
|
import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -7,7 +7,6 @@ import pytest
|
||||||
import torch
|
import torch
|
||||||
|
|
||||||
from grouped_gemm.interface import (
|
from grouped_gemm.interface import (
|
||||||
grouped_gemm,
|
|
||||||
grouped_gemm_dW,
|
grouped_gemm_dW,
|
||||||
grouped_gemm_dX,
|
grouped_gemm_dX,
|
||||||
grouped_gemm_forward,
|
grouped_gemm_forward,
|
||||||
|
|
@ -582,7 +581,6 @@ def _test_grouped_gemm_backward_dX(
|
||||||
kernel_config_bwd_dW = KernelConfigBackward_dW()
|
kernel_config_bwd_dW = KernelConfigBackward_dW()
|
||||||
else:
|
else:
|
||||||
from grouped_gemm.kernels.backward import (
|
from grouped_gemm.kernels.backward import (
|
||||||
_autotuned_grouped_gemm_dW_kernel,
|
|
||||||
_autotuned_grouped_gemm_dX_kernel,
|
_autotuned_grouped_gemm_dX_kernel,
|
||||||
)
|
)
|
||||||
from grouped_gemm.kernels.forward import (
|
from grouped_gemm.kernels.forward import (
|
||||||
|
|
|
||||||
|
|
@ -2,7 +2,6 @@
|
||||||
# Copyright 2023-present the Unsloth team. All rights reserved.
|
# Copyright 2023-present the Unsloth team. All rights reserved.
|
||||||
|
|
||||||
import argparse
|
import argparse
|
||||||
import sys
|
|
||||||
from contextlib import contextmanager
|
from contextlib import contextmanager
|
||||||
from functools import partial
|
from functools import partial
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -15,7 +15,7 @@
|
||||||
import triton
|
import triton
|
||||||
import triton.language as tl
|
import triton.language as tl
|
||||||
import torch
|
import torch
|
||||||
from .utils import calculate_settings, torch_gpu_device
|
from .utils import torch_gpu_device
|
||||||
|
|
||||||
# signed int32 max is 2**31-1 so num_elements cannot exceed 2**31
|
# signed int32 max is 2**31-1 so num_elements cannot exceed 2**31
|
||||||
NUM_INT32_ELEMENTS = 2**31
|
NUM_INT32_ELEMENTS = 2**31
|
||||||
|
|
|
||||||
|
|
@ -19,18 +19,13 @@ import ctypes
|
||||||
MAX_FUSED_SIZE: int = 65536
|
MAX_FUSED_SIZE: int = 65536
|
||||||
next_power_of_2 = triton.next_power_of_2
|
next_power_of_2 = triton.next_power_of_2
|
||||||
import functools
|
import functools
|
||||||
from typing import Optional
|
|
||||||
|
|
||||||
from ..device_type import (
|
from ..device_type import (
|
||||||
is_hip,
|
is_hip,
|
||||||
get_device_type,
|
|
||||||
DEVICE_TYPE,
|
DEVICE_TYPE,
|
||||||
DEVICE_TYPE_TORCH,
|
|
||||||
DEVICE_COUNT,
|
DEVICE_COUNT,
|
||||||
ALLOW_PREQUANTIZED_MODELS,
|
|
||||||
)
|
)
|
||||||
from .fp8 import weight_dequant, fp8_linear
|
from .fp8 import weight_dequant, fp8_linear
|
||||||
import functools
|
|
||||||
|
|
||||||
# torch.cuda.amp.custom_fwd is deprecated >= 2.4
|
# torch.cuda.amp.custom_fwd is deprecated >= 2.4
|
||||||
import torch
|
import torch
|
||||||
|
|
|
||||||
|
|
@ -86,35 +86,26 @@ from typing import Union, Optional, List, Any, Callable, Tuple, Iterator
|
||||||
from platform import system as platform_system
|
from platform import system as platform_system
|
||||||
|
|
||||||
platform_system = platform_system()
|
platform_system = platform_system()
|
||||||
import numpy as np
|
|
||||||
import contextlib
|
|
||||||
import re
|
import re
|
||||||
from dataclasses import dataclass, field
|
from dataclasses import dataclass, field
|
||||||
import functools
|
import functools
|
||||||
import textwrap
|
import textwrap
|
||||||
import logging
|
import logging
|
||||||
import warnings, subprocess, inspect, psutil, os, math
|
import warnings
|
||||||
|
import inspect
|
||||||
|
import psutil
|
||||||
|
import os
|
||||||
from unsloth_zoo.utils import Version, get_quant_type
|
from unsloth_zoo.utils import Version, get_quant_type
|
||||||
from importlib.metadata import version as importlib_version
|
from importlib.metadata import version as importlib_version
|
||||||
from ..device_type import (
|
from ..device_type import (
|
||||||
is_hip,
|
|
||||||
get_device_type,
|
|
||||||
DEVICE_TYPE,
|
DEVICE_TYPE,
|
||||||
DEVICE_TYPE_TORCH,
|
|
||||||
DEVICE_COUNT,
|
DEVICE_COUNT,
|
||||||
ALLOW_PREQUANTIZED_MODELS,
|
|
||||||
)
|
)
|
||||||
from ..import_fixes import UNSLOTH_ENABLE_LOGGING
|
from ..import_fixes import UNSLOTH_ENABLE_LOGGING
|
||||||
from unsloth_zoo.log import logger
|
from unsloth_zoo.log import logger
|
||||||
from unsloth_zoo.tokenizer_utils import (
|
from unsloth_zoo.tokenizer_utils import (
|
||||||
patch_tokenizer as _patch_tokenizer,
|
patch_tokenizer as _patch_tokenizer,
|
||||||
)
|
)
|
||||||
from unsloth_zoo.rl_environments import (
|
|
||||||
check_python_modules,
|
|
||||||
create_locked_down_function,
|
|
||||||
execute_with_time_limit,
|
|
||||||
Benchmarker,
|
|
||||||
)
|
|
||||||
from unsloth_zoo.patching_utils import (
|
from unsloth_zoo.patching_utils import (
|
||||||
patch_compiling_bitsandbytes,
|
patch_compiling_bitsandbytes,
|
||||||
patch_layernorm,
|
patch_layernorm,
|
||||||
|
|
@ -127,8 +118,6 @@ from unsloth_zoo.gradient_checkpointing import (
|
||||||
unsloth_offloaded_gradient_checkpoint,
|
unsloth_offloaded_gradient_checkpoint,
|
||||||
patch_unsloth_gradient_checkpointing,
|
patch_unsloth_gradient_checkpointing,
|
||||||
unpatch_unsloth_gradient_checkpointing,
|
unpatch_unsloth_gradient_checkpointing,
|
||||||
Unsloth_Gradient_Checkpointer,
|
|
||||||
unsloth_gradient_checkpoint,
|
|
||||||
patch_gradient_checkpointing,
|
patch_gradient_checkpointing,
|
||||||
unpatch_gradient_checkpointing,
|
unpatch_gradient_checkpointing,
|
||||||
patch_unsloth_smart_gradient_checkpointing,
|
patch_unsloth_smart_gradient_checkpointing,
|
||||||
|
|
@ -313,7 +302,7 @@ class HideLoggingMessage(logging.Filter):
|
||||||
self.text = text
|
self.text = text
|
||||||
|
|
||||||
def filter(self, x):
|
def filter(self, x):
|
||||||
return not (self.text in x.getMessage())
|
return self.text not in x.getMessage()
|
||||||
|
|
||||||
|
|
||||||
# Replace warning messages (analogous to HideLoggingMessage but for warnings.warn)
|
# Replace warning messages (analogous to HideLoggingMessage but for warnings.warn)
|
||||||
|
|
@ -737,21 +726,21 @@ def patch_mistral_nemo_config(config):
|
||||||
try:
|
try:
|
||||||
# Some Config files use layer_type_validation
|
# Some Config files use layer_type_validation
|
||||||
# for eg Gemma-2, so we must import it to stop errors.
|
# for eg Gemma-2, so we must import it to stop errors.
|
||||||
from transformers.configuration_utils import layer_type_validation
|
pass
|
||||||
except:
|
except:
|
||||||
pass
|
pass
|
||||||
|
|
||||||
try:
|
try:
|
||||||
# Transformers 5.0+ uses RotaryEmbeddingConfigMixin as a base class for configs
|
# Transformers 5.0+ uses RotaryEmbeddingConfigMixin as a base class for configs
|
||||||
from transformers.modeling_rope_utils import RotaryEmbeddingConfigMixin
|
pass
|
||||||
except:
|
except:
|
||||||
pass
|
pass
|
||||||
from transformers import __version__ as transformers_version
|
from transformers import __version__ as transformers_version
|
||||||
|
|
||||||
try:
|
try:
|
||||||
from transformers import PreTrainedConfig
|
pass
|
||||||
except:
|
except:
|
||||||
from transformers import PretrainedConfig
|
pass
|
||||||
|
|
||||||
model_architectures = [
|
model_architectures = [
|
||||||
"llama",
|
"llama",
|
||||||
|
|
@ -850,7 +839,7 @@ from transformers.utils import is_openai_available
|
||||||
|
|
||||||
if is_openai_available():
|
if is_openai_available():
|
||||||
try:
|
try:
|
||||||
from openai import OpenAI
|
pass
|
||||||
except:
|
except:
|
||||||
print("Unsloth: OpenAI failed to import - ignoring for now.")
|
print("Unsloth: OpenAI failed to import - ignoring for now.")
|
||||||
import transformers.utils
|
import transformers.utils
|
||||||
|
|
@ -862,9 +851,7 @@ if is_openai_available():
|
||||||
|
|
||||||
# =============================================
|
# =============================================
|
||||||
# Get Flash Attention v2 if Ampere (RTX 30xx, A100)
|
# Get Flash Attention v2 if Ampere (RTX 30xx, A100)
|
||||||
import bitsandbytes as bnb
|
|
||||||
|
|
||||||
from transformers import AutoTokenizer
|
|
||||||
from transformers.utils.import_utils import _is_package_available
|
from transformers.utils.import_utils import _is_package_available
|
||||||
|
|
||||||
SUPPORTS_BFLOAT16 = False
|
SUPPORTS_BFLOAT16 = False
|
||||||
|
|
@ -882,9 +869,9 @@ if DEVICE_TYPE == "cuda":
|
||||||
try:
|
try:
|
||||||
try:
|
try:
|
||||||
# See https://github.com/unslothai/unsloth/issues/1437
|
# See https://github.com/unslothai/unsloth/issues/1437
|
||||||
from flash_attn.flash_attn_interface import flash_attn_gpu
|
pass
|
||||||
except:
|
except:
|
||||||
from flash_attn.flash_attn_interface import flash_attn_cuda
|
pass
|
||||||
HAS_FLASH_ATTENTION = True
|
HAS_FLASH_ATTENTION = True
|
||||||
|
|
||||||
# Also check for softcapping
|
# Also check for softcapping
|
||||||
|
|
@ -931,9 +918,9 @@ elif DEVICE_TYPE == "hip":
|
||||||
try:
|
try:
|
||||||
try:
|
try:
|
||||||
# See https://github.com/unslothai/unsloth/issues/1437
|
# See https://github.com/unslothai/unsloth/issues/1437
|
||||||
from flash_attn.flash_attn_interface import flash_attn_gpu
|
pass
|
||||||
except:
|
except:
|
||||||
from flash_attn.flash_attn_interface import flash_attn_cuda
|
pass
|
||||||
HAS_FLASH_ATTENTION = True
|
HAS_FLASH_ATTENTION = True
|
||||||
|
|
||||||
# Also check for softcapping
|
# Also check for softcapping
|
||||||
|
|
@ -1286,7 +1273,7 @@ USE_MODELSCOPE = os.environ.get("UNSLOTH_USE_MODELSCOPE", "0") == "1"
|
||||||
if USE_MODELSCOPE:
|
if USE_MODELSCOPE:
|
||||||
if importlib.util.find_spec("modelscope") is None:
|
if importlib.util.find_spec("modelscope") is None:
|
||||||
raise ImportError(
|
raise ImportError(
|
||||||
f"You are using the modelscope hub, please install modelscope by `pip install modelscope -U`"
|
"You are using the modelscope hub, please install modelscope by `pip install modelscope -U`"
|
||||||
)
|
)
|
||||||
|
|
||||||
import socket
|
import socket
|
||||||
|
|
@ -1304,13 +1291,10 @@ def has_internet(host = "8.8.8.8", port = 53, timeout = 3):
|
||||||
return True
|
return True
|
||||||
finally:
|
finally:
|
||||||
sock.close()
|
sock.close()
|
||||||
except socket.error as ex:
|
except socket.error:
|
||||||
return False
|
return False
|
||||||
|
|
||||||
|
|
||||||
import psutil
|
|
||||||
|
|
||||||
|
|
||||||
def _get_statistics(statistics = None, force_download = True):
|
def _get_statistics(statistics = None, force_download = True):
|
||||||
# We log some basic stats about which environment is being used.
|
# We log some basic stats about which environment is being used.
|
||||||
# We simply download a README.md file from HF - all data is made public.
|
# We simply download a README.md file from HF - all data is made public.
|
||||||
|
|
@ -1477,7 +1461,6 @@ def get_statistics(local_files_only = False):
|
||||||
# Fixes Bitsandbytes to remove missing warnings
|
# Fixes Bitsandbytes to remove missing warnings
|
||||||
from transformers.utils.quantization_config import (
|
from transformers.utils.quantization_config import (
|
||||||
BitsAndBytesConfig,
|
BitsAndBytesConfig,
|
||||||
QuantizationMethod,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
BitsAndBytesConfig__init__ = inspect.getsource(BitsAndBytesConfig.__init__)
|
BitsAndBytesConfig__init__ = inspect.getsource(BitsAndBytesConfig.__init__)
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
from unsloth_zoo.utils import _get_dtype, Version
|
from unsloth_zoo.utils import _get_dtype, Version
|
||||||
from ..utils.packing import get_packed_info_from_kwargs
|
from ..utils.packing import get_packed_info_from_kwargs
|
||||||
|
|
@ -30,9 +29,6 @@ try:
|
||||||
CohereDecoderLayer,
|
CohereDecoderLayer,
|
||||||
CohereModel,
|
CohereModel,
|
||||||
CohereForCausalLM,
|
CohereForCausalLM,
|
||||||
CohereRotaryEmbedding,
|
|
||||||
apply_rotary_pos_emb,
|
|
||||||
repeat_kv,
|
|
||||||
)
|
)
|
||||||
except:
|
except:
|
||||||
transformers_version = Version(transformers_version)
|
transformers_version = Version(transformers_version)
|
||||||
|
|
|
||||||
|
|
@ -13,8 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.utils import Version, _get_dtype
|
from unsloth_zoo.utils import Version, _get_dtype
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
from ..utils.packing import get_packed_info_from_kwargs
|
from ..utils.packing import get_packed_info_from_kwargs
|
||||||
|
|
@ -28,7 +26,6 @@ from ..utils.attention_dispatch import (
|
||||||
from .llama import (
|
from .llama import (
|
||||||
LlamaRotaryEmbedding,
|
LlamaRotaryEmbedding,
|
||||||
LlamaLinearScalingRotaryEmbedding,
|
LlamaLinearScalingRotaryEmbedding,
|
||||||
_LlamaModel_fast_forward_inference,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
try:
|
try:
|
||||||
|
|
@ -37,7 +34,6 @@ try:
|
||||||
FalconH1DecoderLayer,
|
FalconH1DecoderLayer,
|
||||||
FalconH1Model,
|
FalconH1Model,
|
||||||
FalconH1ForCausalLM,
|
FalconH1ForCausalLM,
|
||||||
FalconHybridMambaAttentionDynamicCache,
|
|
||||||
)
|
)
|
||||||
except:
|
except:
|
||||||
from transformers import __version__ as transformers_version
|
from transformers import __version__ as transformers_version
|
||||||
|
|
|
||||||
|
|
@ -14,14 +14,8 @@
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
from .llama import _get_rope_theta
|
from .llama import _get_rope_theta
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.utils import _get_dtype, Version
|
from unsloth_zoo.utils import _get_dtype, Version
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
from ..utils.packing import (
|
|
||||||
build_sdpa_packed_attention_mask,
|
|
||||||
build_xformers_block_causal_mask,
|
|
||||||
get_packed_info_from_kwargs,
|
|
||||||
)
|
|
||||||
import math
|
import math
|
||||||
|
|
||||||
try:
|
try:
|
||||||
|
|
@ -30,9 +24,6 @@ try:
|
||||||
GemmaDecoderLayer,
|
GemmaDecoderLayer,
|
||||||
GemmaModel,
|
GemmaModel,
|
||||||
GemmaForCausalLM,
|
GemmaForCausalLM,
|
||||||
GemmaRotaryEmbedding,
|
|
||||||
apply_rotary_pos_emb,
|
|
||||||
repeat_kv,
|
|
||||||
)
|
)
|
||||||
except:
|
except:
|
||||||
transformers_version = Version(transformers_version)
|
transformers_version = Version(transformers_version)
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.utils import _get_dtype, Version
|
from unsloth_zoo.utils import _get_dtype, Version
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
from ..utils.packing import get_packed_info_from_kwargs
|
from ..utils.packing import get_packed_info_from_kwargs
|
||||||
|
|
@ -22,7 +21,6 @@ from ..utils.attention_dispatch import (
|
||||||
AttentionContext,
|
AttentionContext,
|
||||||
run_attention,
|
run_attention,
|
||||||
select_attention_backend,
|
select_attention_backend,
|
||||||
SDPA,
|
|
||||||
)
|
)
|
||||||
from .gemma import (
|
from .gemma import (
|
||||||
GemmaFixedRotaryEmbedding,
|
GemmaFixedRotaryEmbedding,
|
||||||
|
|
@ -36,9 +34,6 @@ try:
|
||||||
Gemma2DecoderLayer,
|
Gemma2DecoderLayer,
|
||||||
Gemma2Model,
|
Gemma2Model,
|
||||||
Gemma2ForCausalLM,
|
Gemma2ForCausalLM,
|
||||||
Gemma2RotaryEmbedding,
|
|
||||||
apply_rotary_pos_emb,
|
|
||||||
repeat_kv,
|
|
||||||
)
|
)
|
||||||
except:
|
except:
|
||||||
transformers_version = Version(transformers_version)
|
transformers_version = Version(transformers_version)
|
||||||
|
|
@ -65,7 +60,7 @@ except:
|
||||||
Gemma2FlashAttention2 = Gemma2Attention
|
Gemma2FlashAttention2 = Gemma2Attention
|
||||||
|
|
||||||
if HAS_FLASH_ATTENTION_SOFTCAPPING:
|
if HAS_FLASH_ATTENTION_SOFTCAPPING:
|
||||||
from flash_attn import flash_attn_func
|
pass
|
||||||
|
|
||||||
|
|
||||||
# Logit softcapping
|
# Logit softcapping
|
||||||
|
|
|
||||||
|
|
@ -25,20 +25,11 @@ Key architecture differences from Qwen3 MoE:
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
import os
|
||||||
from ._utils import __version__
|
|
||||||
from .llama import (
|
from .llama import (
|
||||||
LlamaRotaryEmbedding,
|
|
||||||
LlamaLinearScalingRotaryEmbedding,
|
|
||||||
fix_prepare_inputs_for_generation,
|
|
||||||
fast_rms_layernorm_inference,
|
fast_rms_layernorm_inference,
|
||||||
fast_swiglu_inference,
|
fast_swiglu_inference,
|
||||||
LlamaModel_fast_forward,
|
|
||||||
LlamaModel_fast_forward_inference,
|
|
||||||
CausalLM_fast_forward,
|
|
||||||
PeftModel_fast_forward,
|
|
||||||
)
|
)
|
||||||
import torch
|
import torch
|
||||||
import torch.nn.functional as F
|
|
||||||
from typing import Optional, Tuple
|
from typing import Optional, Tuple
|
||||||
from ..kernels import fast_rms_layernorm
|
from ..kernels import fast_rms_layernorm
|
||||||
|
|
||||||
|
|
|
||||||
|
|
@ -13,7 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
|
||||||
from ._utils import __version__
|
from ._utils import __version__
|
||||||
from unsloth_zoo.utils import _get_dtype, Version
|
from unsloth_zoo.utils import _get_dtype, Version
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
|
|
@ -265,8 +264,6 @@ def GraniteDecoderLayer_fast_forward(
|
||||||
return outputs
|
return outputs
|
||||||
|
|
||||||
|
|
||||||
from math import sqrt as math_sqrt
|
|
||||||
|
|
||||||
KV_CACHE_INCREMENT = 256 # KV Cache update size
|
KV_CACHE_INCREMENT = 256 # KV Cache update size
|
||||||
torch_nn_functional_softmax = torch.nn.functional.softmax
|
torch_nn_functional_softmax = torch.nn.functional.softmax
|
||||||
torch_matmul = torch.matmul
|
torch_matmul = torch.matmul
|
||||||
|
|
@ -285,7 +282,7 @@ def GraniteAttention_fast_forward_inference(
|
||||||
):
|
):
|
||||||
assert (
|
assert (
|
||||||
position_embeddings is not None
|
position_embeddings is not None
|
||||||
), f"Granite model requires position embeddings to be specified"
|
), "Granite model requires position embeddings to be specified"
|
||||||
|
|
||||||
Xn = hidden_states
|
Xn = hidden_states
|
||||||
bsz, _, hd = hidden_states.size()
|
bsz, _, hd = hidden_states.size()
|
||||||
|
|
|
||||||
|
|
@ -26,7 +26,6 @@ from ._utils import (
|
||||||
_get_inference_mode_context_manager,
|
_get_inference_mode_context_manager,
|
||||||
_prepare_model_for_qat,
|
_prepare_model_for_qat,
|
||||||
is_bfloat16_supported,
|
is_bfloat16_supported,
|
||||||
get_quant_type,
|
|
||||||
)
|
)
|
||||||
from .loader_utils import _get_fp8_mode_and_check_settings
|
from .loader_utils import _get_fp8_mode_and_check_settings
|
||||||
from ..utils.packing import (
|
from ..utils.packing import (
|
||||||
|
|
@ -50,12 +49,9 @@ from unsloth_zoo.hf_utils import (
|
||||||
)
|
)
|
||||||
from unsloth_zoo.peft_utils import SKIP_QUANTIZATION_MODULES
|
from unsloth_zoo.peft_utils import SKIP_QUANTIZATION_MODULES
|
||||||
from ..device_type import (
|
from ..device_type import (
|
||||||
is_hip,
|
|
||||||
get_device_type,
|
|
||||||
DEVICE_TYPE,
|
DEVICE_TYPE,
|
||||||
DEVICE_TYPE_TORCH,
|
DEVICE_TYPE_TORCH,
|
||||||
DEVICE_COUNT,
|
DEVICE_COUNT,
|
||||||
ALLOW_PREQUANTIZED_MODELS,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
transformers_version = Version(transformers_version)
|
transformers_version = Version(transformers_version)
|
||||||
|
|
@ -97,7 +93,6 @@ except:
|
||||||
LlamaFlashAttention2 = LlamaAttention
|
LlamaFlashAttention2 = LlamaAttention
|
||||||
|
|
||||||
from transformers import (
|
from transformers import (
|
||||||
AutoTokenizer,
|
|
||||||
AutoModelForCausalLM,
|
AutoModelForCausalLM,
|
||||||
AutoModelForSequenceClassification,
|
AutoModelForSequenceClassification,
|
||||||
BitsAndBytesConfig,
|
BitsAndBytesConfig,
|
||||||
|
|
@ -108,14 +103,16 @@ from transformers import set_seed as transformers_set_seed
|
||||||
from peft import LoraConfig, TaskType, get_peft_model as _get_peft_model
|
from peft import LoraConfig, TaskType, get_peft_model as _get_peft_model
|
||||||
from peft import PeftModelForCausalLM, PeftModelForSequenceClassification
|
from peft import PeftModelForCausalLM, PeftModelForSequenceClassification
|
||||||
from ..save import patch_saving_functions
|
from ..save import patch_saving_functions
|
||||||
import re, os, inspect, math, sys
|
import re
|
||||||
|
import os
|
||||||
|
import inspect
|
||||||
import types
|
import types
|
||||||
|
|
||||||
try:
|
try:
|
||||||
from huggingface_hub.utils import get_token
|
pass
|
||||||
except:
|
except:
|
||||||
# Old HF Hub versions <= 0.0.25
|
# Old HF Hub versions <= 0.0.25
|
||||||
from huggingface_hub.utils._token import get_token
|
pass
|
||||||
from triton import __version__ as triton_version
|
from triton import __version__ as triton_version
|
||||||
|
|
||||||
HAS_XFORMERS = xformers is not None
|
HAS_XFORMERS = xformers is not None
|
||||||
|
|
@ -2983,7 +2980,7 @@ class FastLlamaModel:
|
||||||
try:
|
try:
|
||||||
assert module in accepted_modules
|
assert module in accepted_modules
|
||||||
final_modules.append(module)
|
final_modules.append(module)
|
||||||
except AssertionError as e:
|
except AssertionError:
|
||||||
final_modules.append(module)
|
final_modules.append(module)
|
||||||
print(
|
print(
|
||||||
"Unsloth: You added custom modules, but Unsloth hasn't optimized for this.\n"
|
"Unsloth: You added custom modules, but Unsloth hasn't optimized for this.\n"
|
||||||
|
|
|
||||||
|
|
@ -15,20 +15,17 @@
|
||||||
from ._utils import (
|
from ._utils import (
|
||||||
_prepare_model_for_qat,
|
_prepare_model_for_qat,
|
||||||
is_bfloat16_supported,
|
is_bfloat16_supported,
|
||||||
is_vLLM_available,
|
|
||||||
HAS_FLASH_ATTENTION,
|
HAS_FLASH_ATTENTION,
|
||||||
HAS_FLASH_ATTENTION_SOFTCAPPING,
|
HAS_FLASH_ATTENTION_SOFTCAPPING,
|
||||||
USE_MODELSCOPE,
|
USE_MODELSCOPE,
|
||||||
get_transformers_model_type,
|
get_transformers_model_type,
|
||||||
hf_login,
|
hf_login,
|
||||||
)
|
)
|
||||||
from .granite import FastGraniteModel
|
|
||||||
from .llama import FastLlamaModel, logger
|
from .llama import FastLlamaModel, logger
|
||||||
from .mistral import FastMistralModel
|
from .mistral import FastMistralModel
|
||||||
from .qwen2 import FastQwen2Model
|
from .qwen2 import FastQwen2Model
|
||||||
from .qwen3 import FastQwen3Model
|
from .qwen3 import FastQwen3Model
|
||||||
from .qwen3_moe import FastQwen3MoeModel
|
from .qwen3_moe import FastQwen3MoeModel
|
||||||
from .cohere import FastCohereModel
|
|
||||||
from transformers import AutoConfig
|
from transformers import AutoConfig
|
||||||
from transformers import __version__ as transformers_version
|
from transformers import __version__ as transformers_version
|
||||||
from peft import PeftConfig, PeftModel
|
from peft import PeftConfig, PeftModel
|
||||||
|
|
@ -39,22 +36,20 @@ from .loader_utils import (
|
||||||
get_model_name,
|
get_model_name,
|
||||||
prepare_device_map,
|
prepare_device_map,
|
||||||
)
|
)
|
||||||
import os, contextlib, sys
|
import os
|
||||||
|
import contextlib
|
||||||
|
|
||||||
try:
|
try:
|
||||||
from huggingface_hub import get_token
|
pass
|
||||||
except:
|
except:
|
||||||
try:
|
try:
|
||||||
from huggingface_hub.utils import get_token
|
pass
|
||||||
except:
|
except:
|
||||||
# For older versions of huggingface_hub
|
# For older versions of huggingface_hub
|
||||||
from huggingface_hub.utils._token import get_token
|
pass
|
||||||
from huggingface_hub import HfFileSystem
|
from huggingface_hub import HfFileSystem
|
||||||
import importlib.util
|
import importlib.util
|
||||||
from ..device_type import (
|
from ..device_type import (
|
||||||
is_hip,
|
|
||||||
get_device_type,
|
|
||||||
DEVICE_TYPE,
|
|
||||||
DEVICE_TYPE_TORCH,
|
DEVICE_TYPE_TORCH,
|
||||||
DEVICE_COUNT,
|
DEVICE_COUNT,
|
||||||
ALLOW_PREQUANTIZED_MODELS,
|
ALLOW_PREQUANTIZED_MODELS,
|
||||||
|
|
@ -86,15 +81,12 @@ if SUPPORTS_GEMMA:
|
||||||
if SUPPORTS_GEMMA2:
|
if SUPPORTS_GEMMA2:
|
||||||
from .gemma2 import FastGemma2Model
|
from .gemma2 import FastGemma2Model
|
||||||
if SUPPORTS_FALCON_H1:
|
if SUPPORTS_FALCON_H1:
|
||||||
from .falcon_h1 import FastFalconH1Model
|
pass
|
||||||
import torch
|
import torch
|
||||||
from ._utils import (
|
from ._utils import (
|
||||||
patch_compiling_bitsandbytes,
|
patch_compiling_bitsandbytes,
|
||||||
patch_model_and_tokenizer,
|
|
||||||
prepare_model_for_kbit_training,
|
|
||||||
apply_unsloth_gradient_checkpointing,
|
apply_unsloth_gradient_checkpointing,
|
||||||
patch_compiled_autograd,
|
patch_compiled_autograd,
|
||||||
process_vision_info,
|
|
||||||
unsloth_compile_transformers,
|
unsloth_compile_transformers,
|
||||||
fast_inference_setup,
|
fast_inference_setup,
|
||||||
)
|
)
|
||||||
|
|
@ -806,7 +798,6 @@ class FastLanguageModel(FastLlamaModel):
|
||||||
|
|
||||||
from ..kernels import (
|
from ..kernels import (
|
||||||
patch_loss_functions,
|
patch_loss_functions,
|
||||||
post_patch_loss_function,
|
|
||||||
)
|
)
|
||||||
from .vision import FastBaseModel
|
from .vision import FastBaseModel
|
||||||
from transformers import (
|
from transformers import (
|
||||||
|
|
|
||||||
|
|
@ -16,7 +16,6 @@ from ..device_type import DEVICE_TYPE_TORCH
|
||||||
import importlib
|
import importlib
|
||||||
import os
|
import os
|
||||||
import torch
|
import torch
|
||||||
import re
|
|
||||||
import tempfile
|
import tempfile
|
||||||
from typing import Union
|
from typing import Union
|
||||||
from .mapper import (
|
from .mapper import (
|
||||||
|
|
|
||||||
File diff suppressed because it is too large
Load diff
|
|
@ -14,7 +14,6 @@
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
import os
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.utils import _get_dtype
|
from unsloth_zoo.utils import _get_dtype
|
||||||
from unsloth_zoo.hf_utils import dtype_from_config
|
from unsloth_zoo.hf_utils import dtype_from_config
|
||||||
from ..utils.packing import (
|
from ..utils.packing import (
|
||||||
|
|
@ -48,7 +47,6 @@ try:
|
||||||
except:
|
except:
|
||||||
MistralSdpaAttention = MistralAttention
|
MistralSdpaAttention = MistralAttention
|
||||||
MistralFlashAttention2 = MistralAttention
|
MistralFlashAttention2 = MistralAttention
|
||||||
from unsloth_zoo.utils import Version, _get_dtype
|
|
||||||
|
|
||||||
|
|
||||||
def MistralAttention_fast_forward(
|
def MistralAttention_fast_forward(
|
||||||
|
|
|
||||||
|
|
@ -13,9 +13,7 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
from unsloth_zoo.utils import Version
|
||||||
from ._utils import __version__
|
|
||||||
from unsloth_zoo.utils import Version, _get_dtype
|
|
||||||
from ..utils.packing import get_packed_info_from_kwargs
|
from ..utils.packing import get_packed_info_from_kwargs
|
||||||
from ..utils.attention_dispatch import (
|
from ..utils.attention_dispatch import (
|
||||||
AttentionConfig,
|
AttentionConfig,
|
||||||
|
|
@ -48,9 +46,6 @@ except:
|
||||||
f'Try `pip install --upgrade "transformers>=4.50.3"`\n'
|
f'Try `pip install --upgrade "transformers>=4.50.3"`\n'
|
||||||
f"to obtain the latest transformers build, then restart this session."
|
f"to obtain the latest transformers build, then restart this session."
|
||||||
)
|
)
|
||||||
from transformers.modeling_attn_mask_utils import (
|
|
||||||
_prepare_4d_causal_attention_mask_for_sdpa,
|
|
||||||
)
|
|
||||||
|
|
||||||
# For Pytorch 2.1.1
|
# For Pytorch 2.1.1
|
||||||
try:
|
try:
|
||||||
|
|
|
||||||
|
|
@ -13,8 +13,6 @@
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from .llama import *
|
from .llama import *
|
||||||
import os
|
|
||||||
from ._utils import __version__
|
|
||||||
from .llama import (
|
from .llama import (
|
||||||
LlamaRotaryEmbedding,
|
LlamaRotaryEmbedding,
|
||||||
LlamaLinearScalingRotaryEmbedding,
|
LlamaLinearScalingRotaryEmbedding,
|
||||||
|
|
@ -43,7 +41,6 @@ from transformers.models.qwen3_moe.modeling_qwen3_moe import (
|
||||||
# Qwen3SdpaAttention = Qwen3Attention
|
# Qwen3SdpaAttention = Qwen3Attention
|
||||||
# Qwen3FlashAttention2 = Qwen3Attention
|
# Qwen3FlashAttention2 = Qwen3Attention
|
||||||
# pass
|
# pass
|
||||||
from unsloth_zoo.utils import Version, _get_dtype
|
|
||||||
|
|
||||||
|
|
||||||
torch_nn_functional_softmax = torch.nn.functional.softmax
|
torch_nn_functional_softmax = torch.nn.functional.softmax
|
||||||
|
|
|
||||||
Some files were not shown because too many files have changed in this diff Show more
Loading…
Add table
Add a link
Reference in a new issue