huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-12 12:54:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

Common arch list.

Browse Source
This commit is contained in:
Nicolas Patry 2024-08-16 19:02:22 +02:00
parent 9190b0c82c
commit f18262b16f
No known key found for this signature in database
GPG Key ID: 64AF4752B2967863
1 changed files with 6 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
Dockerfile
View File

@ -88,6 +88,7 @@ RUN case ${TARGETPLATFORM} in \
FROM pytorch-install AS kernel-builder FROM pytorch-install AS kernel-builder
ARG MAX_JOBS=8 ARG MAX_JOBS=8
ENV TORCH_CUDA_ARCH_LIST="8.0;8.6;9.0+PTX"
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
ninja-build cmake \ ninja-build cmake \
@ -118,7 +119,7 @@ FROM kernel-builder AS exllama-kernels-builder
WORKDIR /usr/src WORKDIR /usr/src
COPY server/exllama_kernels/ . COPY server/exllama_kernels/ .
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" python setup.py build RUN python setup.py build
# Build Transformers exllama kernels # Build Transformers exllama kernels
FROM kernel-builder AS exllamav2-kernels-builder FROM kernel-builder AS exllamav2-kernels-builder
@ -126,28 +127,28 @@ WORKDIR /usr/src
COPY server/Makefile-exllamav2/ Makefile COPY server/Makefile-exllamav2/ Makefile
# Build specific version of transformers # Build specific version of transformers
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-exllamav2 RUN make build-exllamav2
# Build Transformers awq kernels # Build Transformers awq kernels
FROM kernel-builder AS awq-kernels-builder FROM kernel-builder AS awq-kernels-builder
WORKDIR /usr/src WORKDIR /usr/src
COPY server/Makefile-awq Makefile COPY server/Makefile-awq Makefile
# Build specific version of transformers # Build specific version of transformers
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-awq RUN make build-awq
# Build eetq kernels # Build eetq kernels
FROM kernel-builder AS eetq-kernels-builder FROM kernel-builder AS eetq-kernels-builder
WORKDIR /usr/src WORKDIR /usr/src
COPY server/Makefile-eetq Makefile COPY server/Makefile-eetq Makefile
# Build specific version of transformers # Build specific version of transformers
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-eetq RUN make build-eetq
# Build Lorax Punica kernels # Build Lorax Punica kernels
FROM kernel-builder AS lorax-punica-builder FROM kernel-builder AS lorax-punica-builder
WORKDIR /usr/src WORKDIR /usr/src
COPY server/Makefile-lorax-punica Makefile COPY server/Makefile-lorax-punica Makefile
# Build specific version of transformers # Build specific version of transformers
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-lorax-punica RUN make build-lorax-punica
# Build Transformers CUDA kernels # Build Transformers CUDA kernels
FROM kernel-builder AS custom-kernels-builder FROM kernel-builder AS custom-kernels-builder