huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-11 20:34:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

Making easier local install.

Browse Source
This commit is contained in:
Nicolas Patry 2024-06-04 10:15:30 +00:00
parent f3c4d06bae
commit 4fd7c64793
4 changed files with 29 additions and 50 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
server/Makefile
View File

@ -10,7 +10,7 @@ unit-tests:
gen-server:
# Compile protos
pip install grpcio-tools==1.64.0 mypy-protobuf==3.6.0 'types-protobuf' --no-cache-dir
pip install grpcio-tools==1.62.2 mypy-protobuf==3.6.0 'types-protobuf' --no-cache-dir
mkdir text_generation_server/pb || true
python -m grpc_tools.protoc -I../proto --python_out=text_generation_server/pb \
--grpc_python_out=text_generation_server/pb --mypy_out=text_generation_server/pb ../proto/generate.proto

21
server/Makefile-flash-att
View File

@ -1,16 +1,9 @@
flash_att_commit := 3a9bfd076f98746c73362328958dbc68d145fbec
flash-attention:
# Clone flash attention
pip install -U packaging ninja --no-cache-dir
git clone https://github.com/HazyResearch/flash-attention.git
build-flash-attention: flash-attention
cd flash-attention && git fetch && git checkout $(flash_att_commit)
cd flash-attention && MAX_JOBS=8 python setup.py build
cd flash-attention/csrc/rotary && python setup.py build
cd flash-attention/csrc/layer_norm && python setup.py build
install-flash-attention: build-flash-attention
pip uninstall flash_attn rotary_emb dropout_layer_norm -y || true
cd flash-attention && python setup.py install && cd csrc/layer_norm && python setup.py install && cd ../rotary && python setup.py install
install-flash-attention:
if [ ! -d 'flash-attention' ]; then \
pip install -U packaging ninja --no-cache-dir && \
git clone https://github.com/HazyResearch/flash-attention.git && \
cd flash-attention && git fetch && git checkout $(flash_att_commit) && \
MAX_JOBS=8 python setup.py install && cd csrc/layer_norm && python setup.py install && cd ../rotary && python setup.py install; \
fi

17
server/Makefile-flash-att-v2
View File

@ -4,15 +4,10 @@ flash_att_v2_commit_rocm := 2554f490101742ccdc56620a938f847f61754be6
install-flash-attention-v2-cuda:
pip install flash-attn==$(flash_att_v2_commit_cuda)
flash-attention-v2-rocm:
# Clone flash attention
pip install -U packaging ninja --no-cache-dir
git clone https://github.com/ROCm/flash-attention.git flash-attention-v2
build-flash-attention-v2-rocm: flash-attention-v2-rocm
cd flash-attention-v2 && git fetch && git checkout $(flash_att_v2_commit_rocm)
cd flash-attention-v2 && git submodule update --init --recursive
cd flash-attention-v2 && GPU_ARCHS="gfx90a;gfx942" PYTORCH_ROCM_ARCH="gfx90a;gfx942" python setup.py build
install-flash-attention-v2-rocm: build-flash-attention-v2-rocm
cd flash-attention-v2 && git submodule update --init --recursive && python setup.py install
if [ ! -d 'flash-attention-v2' ]; then \
pip install -U packaging ninja --no-cache-dir && \
git clone https://github.com/ROCm/flash-attention.git flash-attention-v2 && \
cd flash-attention-v2 && git fetch && git checkout $(flash_att_v2_commit_rocm) && \
git submodule update --init --recursive && GPU_ARCHS="gfx90a;gfx942" PYTORCH_ROCM_ARCH="gfx90a;gfx942" python setup.py install; \
fi

39
server/Makefile-vllm
View File

@ -1,25 +1,16 @@
vllm-cuda:
# Clone vllm
pip install -U ninja packaging --no-cache-dir
git clone https://github.com/Narsil/vllm.git vllm
install-vllm-cuda:
if [ ! -d 'vllm' ]; then \
pip install -U ninja packaging --no-cache-dir && \
git clone https://github.com/Narsil/vllm.git vllm &&\
cd vllm && \
git fetch && git checkout b5dfc61db88a81069e45b44f7cc99bd9e62a60fa &&\
pip install -e .; \
fi
build-vllm-cuda: vllm-cuda
cd vllm && git fetch && git checkout b5dfc61db88a81069e45b44f7cc99bd9e62a60fa
cd vllm && python setup.py build
install-vllm-cuda: build-vllm-cuda
pip uninstall vllm -y || true
cd vllm && python setup.py install
vllm-rocm:
# Clone vllm
pip install -U ninja packaging --no-cache-dir
git clone https://github.com/fxmarty/rocm-vllm.git vllm
build-vllm-rocm: vllm-rocm
cd vllm && git fetch && git checkout ca6913b3c2ffacdcb7d15e914dc34adbc6c89479
cd vllm && PYTORCH_ROCM_ARCH="gfx90a;gfx942" python setup.py install
install-vllm-rocm: build-vllm-rocm
pip uninstall vllm -y || true
cd vllm && python setup.py install
install-vllm-rocm:
if [ ! -d 'vllm' ]; then \
pip install -U ninja packaging --no-cache-dir && \
git clone https://github.com/fxmarty/rocm-vllm.git vllm && \
cd vllm && git fetch && git checkout ca6913b3c2ffacdcb7d15e914dc34adbc6c89479 && \
PYTORCH_ROCM_ARCH="gfx90a;gfx942" pip install -e .; \
fi