Moving those deps directly into pyproject.

huggingface · Feb 6, 2025 · 98c1496 · 98c1496
1 parent 1e5d27a
commit 98c1496
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 17 deletions.
diff --git a/Dockerfile_intel b/Dockerfile_intel
@@ -105,23 +105,16 @@ COPY server/Makefile server/Makefile
 ENV UV_SYSTEM_PYTHON=1
 RUN cd server && \
     pip install -U pip uv && \
-	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --no-install-project && \
+	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --extra intel-xpu --no-install-project && \
     . ./.venv/bin/activate && \
     make gen-server-raw
 
 RUN cd server && \
-	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines && \
+	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --extra intel-xpu && \
     . ./.venv/bin/activate && \
     pwd && \
     text-generation-server --help
 
-RUN . ./server/.venv/bin/activate && pip install https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torch-2.5.0a0%2Bgite84e33f-cp311-cp311-linux_x86_64.whl --no-cache-dir
-RUN . ./server/.venv/bin/activate && pip install https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torchaudio-2.5.0a0%2B56bc006-cp311-cp311-linux_x86_64.whl --no-cache-dir
-RUN . ./server/.venv/bin/activate && pip install https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torchvision-0.20.0a0%2B8e8a208-cp311-cp311-linux_x86_64.whl --no-cache-dir
-RUN . ./server/.venv/bin/activate && pip install https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/oneccl_bind_pt-2.5.0%2Bxpu-cp311-cp311-linux_x86_64.whl --no-cache-dir
-
-RUN . ./server/.venv/bin/activate pip install triton-xpu==3.0.0b2 --no-cache-dir
-
 ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/intel/oneapi/pti/0.9/lib:/opt/conda/lib
 ENV CCL_ZE_IPC_EXCHANGE=sockets
 #ENV TORCH_LLM_ALLREDUCE=1
@@ -194,12 +187,6 @@ RUN case ${TARGETPLATFORM} in \
 RUN conda install -c conda-forge gperftools mkl
 
 
-RUN pip install https://download.pytorch.org/whl/nightly/cpu/torch-2.5.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl
-RUN pip install https://download.pytorch.org/whl/nightly/cpu/torchvision-0.20.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl
-RUN pip install https://download.pytorch.org/whl/nightly/cpu/torchaudio-2.4.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl
-
-RUN pip install triton==3.1.0 py-libnuma
-
 WORKDIR /usr/src
 
 RUN git clone https://github.com/intel/intel-extension-for-pytorch && cd intel-extension-for-pytorch && git checkout b7b552baf64283b594665b8687430fe92990e497
@@ -224,12 +211,12 @@ COPY server/Makefile server/Makefile
 ENV UV_SYSTEM_PYTHON=1
 RUN cd server && \
     pip install -U pip uv && \
-	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --no-install-project && \
+	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --extra intel-cpu --no-install-project && \
     . ./.venv/bin/activate && \
     make gen-server-raw
 
 RUN cd server && \
-	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines && \
+	uv sync --frozen --extra gen --extra accelerate --extra compressed-tensors --extra quantize --extra peft --extra outlines --extra intel-cpu && \
     . ./.venv/bin/activate && \
     pwd && \
     text-generation-server --help

diff --git a/server/pyproject.toml b/server/pyproject.toml
@@ -67,6 +67,21 @@ gen = [
     "grpcio-tools>=1.69.0",
     "mypy-protobuf>=3.6.0",
 ]
+intel-xpu = [
+    "torch-intel-xpu",
+    "torchaudio-intel-xpu",
+    "torchvision-intel-xpu",
+    "oneccl-intel-xpu",
+    "triton-xpu==3.0.0b2"
+]
+intel-cpu = [
+    "torch-intel-cpu",
+    "torchaudio-intel-cpu",
+    "torchvision-intel-cpu",
+    "oneccl-intel-cpu",
+    "triton==3.1.0",
+    "py-libnuma"
+]
 
 [tool.uv.sources]
 attention-kernels.url = "https://github.com/danieldk/attention-kernels/releases/download/v0.2.0.post2/attention_kernels-0.2.0.post2+cu123torch2.5-cp39-abi3-linux_x86_64.whl"
@@ -77,6 +92,13 @@ marlin-kernels = [
   { url = "https://github.com/danieldk/marlin-kernels/releases/download/v0.3.7/marlin_kernels-0.3.7+cu123torch2.5-cp312-cp312-linux_x86_64.whl", marker = "python_version == '3.12'" },
 ]
 moe-kernels.url = "https://github.com/danieldk/moe-kernels/releases/download/v0.8.2/moe_kernels-0.8.2+cu123torch2.5-cp39-abi3-linux_x86_64.whl"
+torch-intel-xpu.url = "https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torch-2.5.0a0%2Bgite84e33f-cp311-cp311-linux_x86_64.whl"
+torchaudio-intel-xpu.url = "https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torchaudio-2.5.0a0%2B56bc006-cp311-cp311-linux_x86_64.whl"
+torchvision-intel-xpu.url = "https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/torchvision-0.20.0a0%2B8e8a208-cp311-cp311-linux_x86_64.whl"
+oneccl-intel-xpu.url = "https://intel-optimized-pytorch.s3.cn-north-1.amazonaws.com.cn/ipex_dev/xpu/oneccl_bind_pt-2.5.0%2Bxpu-cp311-cp311-linux_x86_64.whl"
+torch-intel-cpu.url = "https://download.pytorch.org/whl/nightly/cpu/torch-2.5.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl"
+torchaudio-intel-cpu.url = "https://download.pytorch.org/whl/nightly/cpu/torchaudio-2.4.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl"
+torchvision-intel-cpu.url = "https://download.pytorch.org/whl/nightly/cpu/torchvision-0.20.0.dev20240815%2Bcpu-cp311-cp311-linux_x86_64.whl"
 
 [tool.pytest.ini_options]
 markers = ["private: marks tests as requiring an admin hf token (deselect with '-m \"not private\"')"]