huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-11 12:24:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

enable flashdecoding, prefill chunking and prefix caching

Browse Source 
Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
This commit is contained in:
Wang, Yi A 2024-12-09 20:49:10 -08:00
parent 0a2736e34a
commit fd4f861d2c
1 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
Dockerfile_intel
View File

@ -100,7 +100,6 @@ WORKDIR /usr/src
RUN pip install https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/torch-2.5.0a0%2Bgite84e33f-cp311-cp311-linux_x86_64.whl --no-cache-dir
RUN pip install https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/torchaudio-2.5.0a0%2B56bc006-cp311-cp311-linux_x86_64.whl --no-cache-dir
RUN pip install https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/torchvision-0.20.0a0%2B8e8a208-cp311-cp311-linux_x86_64.whl --no-cache-dir
RUN pip install https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/intel_extension_for_pytorch-2.5.10%2Bgit9d489a8-cp311-cp311-linux_x86_64.whl --no-cache-dir
RUN pip install https://intel-extension-for-pytorch.s3.us-east-1.amazonaws.com/ipex_dev/xpu/oneccl_bind_pt-2.5.0%2Bxpu-cp311-cp311-linux_x86_64.whl --no-cache-dir
RUN pip install triton-xpu==3.0.0b2 --no-cache-dir
@ -119,6 +118,9 @@ ENV CCL_ZE_IPC_EXCHANGE=sockets
#ENV TORCH_LLM_ALLREDUCE=1
#ENV CCL_TOPO_FABRIC_VERTEX_CONNECTION_CHECK=0
RUN git clone https://github.com/intel/intel-extension-for-pytorch && cd intel-extension-for-pytorch && git checkout c3e14caf792ad04824dd921e2fc3f16fca0d462e
RUN cd intel-extension-for-pytorch && git submodule update --init --recursive && USE_AOT_DEVLIST='pvc' BUILD_SEPARATE_OPS=OFF BUILD_WITH_CPU=OFF USE_XETLA=ON python setup.py install && rm -rf /usr/src/intel-extension-for-pytorch
# Install benchmarker
COPY --from=builder /usr/src/target/release-opt/text-generation-benchmark /usr/local/bin/text-generation-benchmark
# Install router
@ -222,9 +224,9 @@ COPY --from=builder /usr/src/target/release-opt/text-generation-router /usr/loca
COPY --from=builder /usr/src/target/release-opt/text-generation-launcher /usr/local/bin/text-generation-launcher
FROM ${PLATFORM} AS final
ENV ATTENTION=paged
ENV PREFIX_CACHING=0
ENV PREFILL_CHUNKING=0
ENV ATTENTION=flashdecoding
ENV PREFIX_CACHING=1
ENV PREFILL_CHUNKING=1
ENV CUDA_GRAPHS=0
ENTRYPOINT ["text-generation-launcher"]
CMD ["--json-output"]