2023-04-16 16:26:47 -06:00
|
|
|
# Rust builder
|
2024-02-12 02:09:29 -07:00
|
|
|
FROM lukemathwalker/cargo-chef:latest-rust-1.75 AS chef
|
2023-03-03 07:07:27 -07:00
|
|
|
WORKDIR /usr/src
|
|
|
|
|
2023-06-08 06:51:52 -06:00
|
|
|
ARG CARGO_REGISTRIES_CRATES_IO_PROTOCOL=sparse
|
|
|
|
|
2023-03-03 07:07:27 -07:00
|
|
|
FROM chef as planner
|
|
|
|
COPY Cargo.toml Cargo.toml
|
|
|
|
COPY rust-toolchain.toml rust-toolchain.toml
|
|
|
|
COPY proto proto
|
2023-05-09 06:39:59 -06:00
|
|
|
COPY benchmark benchmark
|
2023-03-03 07:07:27 -07:00
|
|
|
COPY router router
|
|
|
|
COPY launcher launcher
|
|
|
|
RUN cargo chef prepare --recipe-path recipe.json
|
|
|
|
|
|
|
|
FROM chef AS builder
|
2023-02-13 05:02:45 -07:00
|
|
|
|
2023-04-19 13:36:59 -06:00
|
|
|
ARG GIT_SHA
|
2023-05-02 07:43:19 -06:00
|
|
|
ARG DOCKER_LABEL
|
2023-04-19 13:36:59 -06:00
|
|
|
|
2023-02-13 05:02:45 -07:00
|
|
|
RUN PROTOC_ZIP=protoc-21.12-linux-x86_64.zip && \
|
|
|
|
curl -OL https://github.com/protocolbuffers/protobuf/releases/download/v21.12/$PROTOC_ZIP && \
|
|
|
|
unzip -o $PROTOC_ZIP -d /usr/local bin/protoc && \
|
|
|
|
unzip -o $PROTOC_ZIP -d /usr/local 'include/*' && \
|
|
|
|
rm -f $PROTOC_ZIP
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2023-03-03 07:07:27 -07:00
|
|
|
COPY --from=planner /usr/src/recipe.json recipe.json
|
|
|
|
RUN cargo chef cook --release --recipe-path recipe.json
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2023-03-03 07:07:27 -07:00
|
|
|
COPY Cargo.toml Cargo.toml
|
2022-11-08 09:42:38 -07:00
|
|
|
COPY rust-toolchain.toml rust-toolchain.toml
|
2022-10-14 07:56:21 -06:00
|
|
|
COPY proto proto
|
2023-05-09 06:39:59 -06:00
|
|
|
COPY benchmark benchmark
|
2022-10-14 07:56:21 -06:00
|
|
|
COPY router router
|
2022-10-18 07:19:03 -06:00
|
|
|
COPY launcher launcher
|
2023-03-03 07:07:27 -07:00
|
|
|
RUN cargo build --release
|
2022-10-18 07:19:03 -06:00
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
# Python builder
|
|
|
|
# Adapted from: https://github.com/pytorch/pytorch/blob/master/Dockerfile
|
2024-02-13 14:46:16 -07:00
|
|
|
FROM nvidia/cuda:12.1.0-devel-ubuntu22.04 as pytorch-install
|
2023-04-14 02:12:21 -06:00
|
|
|
|
2023-11-23 05:38:50 -07:00
|
|
|
ARG PYTORCH_VERSION=2.1.1
|
|
|
|
ARG PYTHON_VERSION=3.10
|
2023-08-03 13:54:39 -06:00
|
|
|
# Keep in sync with `server/pyproject.toml
|
2023-11-23 05:38:50 -07:00
|
|
|
ARG CUDA_VERSION=12.1
|
|
|
|
ARG MAMBA_VERSION=23.3.1-1
|
2023-04-16 16:26:47 -06:00
|
|
|
ARG CUDA_CHANNEL=nvidia
|
|
|
|
ARG INSTALL_CHANNEL=pytorch
|
|
|
|
# Automatically set by buildx
|
|
|
|
ARG TARGETPLATFORM
|
|
|
|
|
|
|
|
ENV PATH /opt/conda/bin:$PATH
|
|
|
|
|
|
|
|
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
|
|
|
|
build-essential \
|
|
|
|
ca-certificates \
|
|
|
|
ccache \
|
|
|
|
curl \
|
|
|
|
git && \
|
|
|
|
rm -rf /var/lib/apt/lists/*
|
|
|
|
|
|
|
|
# Install conda
|
|
|
|
# translating Docker's TARGETPLATFORM into mamba arches
|
|
|
|
RUN case ${TARGETPLATFORM} in \
|
|
|
|
"linux/arm64") MAMBA_ARCH=aarch64 ;; \
|
|
|
|
*) MAMBA_ARCH=x86_64 ;; \
|
|
|
|
esac && \
|
|
|
|
curl -fsSL -v -o ~/mambaforge.sh -O "https://github.com/conda-forge/miniforge/releases/download/${MAMBA_VERSION}/Mambaforge-${MAMBA_VERSION}-Linux-${MAMBA_ARCH}.sh"
|
|
|
|
RUN chmod +x ~/mambaforge.sh && \
|
|
|
|
bash ~/mambaforge.sh -b -p /opt/conda && \
|
|
|
|
rm ~/mambaforge.sh
|
|
|
|
|
|
|
|
# Install pytorch
|
|
|
|
# On arm64 we exit with an error code
|
|
|
|
RUN case ${TARGETPLATFORM} in \
|
|
|
|
"linux/arm64") exit 1 ;; \
|
|
|
|
*) /opt/conda/bin/conda update -y conda && \
|
2023-11-23 05:38:50 -07:00
|
|
|
/opt/conda/bin/conda install -c "${INSTALL_CHANNEL}" -c "${CUDA_CHANNEL}" -y "python=${PYTHON_VERSION}" "pytorch=$PYTORCH_VERSION" "pytorch-cuda=$(echo $CUDA_VERSION | cut -d'.' -f 1-2)" ;; \
|
2023-04-16 16:26:47 -06:00
|
|
|
esac && \
|
|
|
|
/opt/conda/bin/conda clean -ya
|
|
|
|
|
|
|
|
# CUDA kernels builder image
|
|
|
|
FROM pytorch-install as kernel-builder
|
|
|
|
|
2023-11-23 05:38:50 -07:00
|
|
|
ARG MAX_JOBS=8
|
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
|
2024-04-10 09:20:25 -06:00
|
|
|
ninja-build cmake \
|
2023-04-16 16:26:47 -06:00
|
|
|
&& rm -rf /var/lib/apt/lists/*
|
|
|
|
|
|
|
|
# Build Flash Attention CUDA kernels
|
|
|
|
FROM kernel-builder as flash-att-builder
|
|
|
|
|
|
|
|
WORKDIR /usr/src
|
|
|
|
|
|
|
|
COPY server/Makefile-flash-att Makefile
|
|
|
|
|
|
|
|
# Build specific version of flash attention
|
|
|
|
RUN make build-flash-attention
|
|
|
|
|
2023-07-18 08:21:18 -06:00
|
|
|
# Build Flash Attention v2 CUDA kernels
|
|
|
|
FROM kernel-builder as flash-att-v2-builder
|
|
|
|
|
|
|
|
WORKDIR /usr/src
|
|
|
|
|
|
|
|
COPY server/Makefile-flash-att-v2 Makefile
|
|
|
|
|
|
|
|
# Build specific version of flash attention v2
|
2023-11-27 06:08:12 -07:00
|
|
|
RUN make build-flash-attention-v2-cuda
|
2023-07-18 08:21:18 -06:00
|
|
|
|
2023-07-21 02:59:00 -06:00
|
|
|
# Build Transformers exllama kernels
|
|
|
|
FROM kernel-builder as exllama-kernels-builder
|
|
|
|
WORKDIR /usr/src
|
|
|
|
COPY server/exllama_kernels/ .
|
2023-11-25 14:38:38 -07:00
|
|
|
|
|
|
|
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" python setup.py build
|
|
|
|
|
|
|
|
# Build Transformers exllama kernels
|
|
|
|
FROM kernel-builder as exllamav2-kernels-builder
|
|
|
|
WORKDIR /usr/src
|
|
|
|
COPY server/exllamav2_kernels/ .
|
|
|
|
|
2023-07-21 02:59:00 -06:00
|
|
|
# Build specific version of transformers
|
|
|
|
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" python setup.py build
|
|
|
|
|
Add AWQ quantization inference support (#1019) (#1054)
# Add AWQ quantization inference support
Fixes
https://github.com/huggingface/text-generation-inference/issues/781
This PR (partially) adds support for AWQ quantization for inference.
More information on AWQ [here](https://arxiv.org/abs/2306.00978). In
general, AWQ is faster and more accurate than GPTQ, which is currently
supported by TGI.
This PR installs 4-bit GEMM custom CUDA kernels released by AWQ authors
(in `requirements.txt`, just one line change).
Quick way to test this PR would be bring up TGI as follows:
```
text-generation-server download-weights abhinavkulkarni/codellama-CodeLlama-7b-Python-hf-w4-g128-awq
text-generation-launcher \
--huggingface-hub-cache ~/.cache/huggingface/hub/ \
--model-id abhinavkulkarni/codellama-CodeLlama-7b-Python-hf-w4-g128-awq \
--trust-remote-code --port 8080 \
--max-input-length 2048 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 \
--quantize awq
```
Please note:
* This PR was tested with FlashAttention v2 and vLLM.
* This PR adds support for AWQ inference, not quantizing the models.
That needs to be done outside of TGI, instructions
[here](https://github.com/mit-han-lab/llm-awq/tree/f084f40bd996f3cf3a0633c1ad7d9d476c318aaa).
* This PR only adds support for `FlashLlama` models for now.
* Multi-GPU setup has not been tested.
* No integration tests have been added so far, will add later if
maintainers are interested in this change.
* This PR can be tested on any of the models released
[here](https://huggingface.co/abhinavkulkarni?sort_models=downloads#models).
Please refer to the linked issue for benchmarks for
[abhinavkulkarni/meta-llama-Llama-2-7b-chat-hf-w4-g128-awq](https://huggingface.co/abhinavkulkarni/meta-llama-Llama-2-7b-chat-hf-w4-g128-awq)
vs
[TheBloke/Llama-2-7b-Chat-GPTQ](https://huggingface.co/TheBloke/Llama-2-7b-Chat-GPTQ).
Please note, AWQ has released faster (and in case of Llama, fused)
kernels for 4-bit GEMM, currently at the top of the `main` branch at
https://github.com/mit-han-lab/llm-awq, but this PR uses an older commit
that has been tested to work. We can switch to latest commit later on.
## Who can review?
@OlivierDehaene OR @Narsil
---------
# What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet
though.
Once merged, your PR is going to appear in the release notes with the
title you set, so make sure it's a great title that fully reflects the
extent of your awesome contribution.
Then, please replace this with a description of the change and which
issue is fixed (if applicable). Please also include relevant motivation
and context. List any dependencies (if any) that are required for this
change.
Once you're done, someone will review your PR shortly (see the section
"Who can review?" below to tag some potential reviewers). They may
suggest changes to make the code even better. If no one reviewed your PR
after a week has passed, don't hesitate to post a new comment
@-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Did you read the [contributor
guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the
[forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes?
Here are the
[documentation
guidelines](https://github.com/huggingface/transformers/tree/main/docs),
and
[here are tips on formatting
docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have
passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the
right person to tag with @
@OlivierDehaene OR @Narsil
-->
---------
Co-authored-by: Abhinav M Kulkarni <abhinavkulkarni@gmail.com>
Co-authored-by: Abhinav Kulkarni <abhinav@concentric.ai>
2023-09-25 07:31:27 -06:00
|
|
|
# Build Transformers awq kernels
|
|
|
|
FROM kernel-builder as awq-kernels-builder
|
|
|
|
WORKDIR /usr/src
|
|
|
|
COPY server/Makefile-awq Makefile
|
|
|
|
# Build specific version of transformers
|
|
|
|
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-awq
|
|
|
|
|
2023-09-29 03:19:06 -06:00
|
|
|
# Build eetq kernels
|
|
|
|
FROM kernel-builder as eetq-kernels-builder
|
|
|
|
WORKDIR /usr/src
|
|
|
|
COPY server/Makefile-eetq Makefile
|
|
|
|
# Build specific version of transformers
|
|
|
|
RUN TORCH_CUDA_ARCH_LIST="8.0;8.6+PTX" make build-eetq
|
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
# Build Transformers CUDA kernels
|
2023-06-08 06:51:52 -06:00
|
|
|
FROM kernel-builder as custom-kernels-builder
|
2023-04-16 16:26:47 -06:00
|
|
|
WORKDIR /usr/src
|
2023-06-08 06:51:52 -06:00
|
|
|
COPY server/custom_kernels/ .
|
2023-04-16 16:26:47 -06:00
|
|
|
# Build specific version of transformers
|
2023-06-08 06:51:52 -06:00
|
|
|
RUN python setup.py build
|
2023-04-16 16:26:47 -06:00
|
|
|
|
2023-06-30 11:09:59 -06:00
|
|
|
# Build vllm CUDA kernels
|
|
|
|
FROM kernel-builder as vllm-builder
|
|
|
|
|
|
|
|
WORKDIR /usr/src
|
|
|
|
|
|
|
|
COPY server/Makefile-vllm Makefile
|
|
|
|
|
|
|
|
# Build specific version of vllm
|
2023-11-27 06:08:12 -07:00
|
|
|
RUN make build-vllm-cuda
|
2023-06-30 11:09:59 -06:00
|
|
|
|
2024-02-08 02:19:45 -07:00
|
|
|
# Build mamba kernels
|
|
|
|
FROM kernel-builder as mamba-builder
|
|
|
|
WORKDIR /usr/src
|
|
|
|
COPY server/Makefile-selective-scan Makefile
|
|
|
|
RUN make build-all
|
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
# Text Generation Inference base image
|
2024-02-12 02:09:29 -07:00
|
|
|
FROM nvidia/cuda:12.1.0-base-ubuntu22.04 as base
|
2023-04-16 16:26:47 -06:00
|
|
|
|
|
|
|
# Conda env
|
|
|
|
ENV PATH=/opt/conda/bin:$PATH \
|
|
|
|
CONDA_PREFIX=/opt/conda
|
|
|
|
|
|
|
|
# Text Generation Inference base env
|
|
|
|
ENV HUGGINGFACE_HUB_CACHE=/data \
|
2023-02-18 06:04:11 -07:00
|
|
|
HF_HUB_ENABLE_HF_TRANSFER=1 \
|
2023-04-16 16:26:47 -06:00
|
|
|
PORT=80
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2023-04-14 11:30:30 -06:00
|
|
|
WORKDIR /usr/src
|
2023-03-24 07:02:14 -06:00
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
|
|
|
|
libssl-dev \
|
|
|
|
ca-certificates \
|
|
|
|
make \
|
2023-09-26 07:23:47 -06:00
|
|
|
curl \
|
2023-04-16 16:26:47 -06:00
|
|
|
&& rm -rf /var/lib/apt/lists/*
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2024-04-10 09:20:25 -06:00
|
|
|
# Copy conda with PyTorch installed
|
|
|
|
COPY --from=pytorch-install /opt/conda /opt/conda
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2023-04-16 16:26:47 -06:00
|
|
|
# Copy build artifacts from flash attention builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=flash-att-builder /usr/src/flash-attention/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
|
|
|
COPY --from=flash-att-builder /usr/src/flash-attention/csrc/layer_norm/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
|
|
|
COPY --from=flash-att-builder /usr/src/flash-attention/csrc/rotary/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-04-14 02:12:21 -06:00
|
|
|
|
2023-07-18 08:21:18 -06:00
|
|
|
# Copy build artifacts from flash attention v2 builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=flash-att-v2-builder /usr/src/flash-attention-v2/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-07-18 08:21:18 -06:00
|
|
|
|
2023-06-30 11:09:59 -06:00
|
|
|
# Copy build artifacts from custom kernels builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=custom-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-07-21 02:59:00 -06:00
|
|
|
# Copy build artifacts from exllama kernels builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=exllama-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-11-25 14:38:38 -07:00
|
|
|
# Copy build artifacts from exllamav2 kernels builder
|
|
|
|
COPY --from=exllamav2-kernels-builder /usr/src/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
Add AWQ quantization inference support (#1019) (#1054)
# Add AWQ quantization inference support
Fixes
https://github.com/huggingface/text-generation-inference/issues/781
This PR (partially) adds support for AWQ quantization for inference.
More information on AWQ [here](https://arxiv.org/abs/2306.00978). In
general, AWQ is faster and more accurate than GPTQ, which is currently
supported by TGI.
This PR installs 4-bit GEMM custom CUDA kernels released by AWQ authors
(in `requirements.txt`, just one line change).
Quick way to test this PR would be bring up TGI as follows:
```
text-generation-server download-weights abhinavkulkarni/codellama-CodeLlama-7b-Python-hf-w4-g128-awq
text-generation-launcher \
--huggingface-hub-cache ~/.cache/huggingface/hub/ \
--model-id abhinavkulkarni/codellama-CodeLlama-7b-Python-hf-w4-g128-awq \
--trust-remote-code --port 8080 \
--max-input-length 2048 --max-total-tokens 4096 --max-batch-prefill-tokens 4096 \
--quantize awq
```
Please note:
* This PR was tested with FlashAttention v2 and vLLM.
* This PR adds support for AWQ inference, not quantizing the models.
That needs to be done outside of TGI, instructions
[here](https://github.com/mit-han-lab/llm-awq/tree/f084f40bd996f3cf3a0633c1ad7d9d476c318aaa).
* This PR only adds support for `FlashLlama` models for now.
* Multi-GPU setup has not been tested.
* No integration tests have been added so far, will add later if
maintainers are interested in this change.
* This PR can be tested on any of the models released
[here](https://huggingface.co/abhinavkulkarni?sort_models=downloads#models).
Please refer to the linked issue for benchmarks for
[abhinavkulkarni/meta-llama-Llama-2-7b-chat-hf-w4-g128-awq](https://huggingface.co/abhinavkulkarni/meta-llama-Llama-2-7b-chat-hf-w4-g128-awq)
vs
[TheBloke/Llama-2-7b-Chat-GPTQ](https://huggingface.co/TheBloke/Llama-2-7b-Chat-GPTQ).
Please note, AWQ has released faster (and in case of Llama, fused)
kernels for 4-bit GEMM, currently at the top of the `main` branch at
https://github.com/mit-han-lab/llm-awq, but this PR uses an older commit
that has been tested to work. We can switch to latest commit later on.
## Who can review?
@OlivierDehaene OR @Narsil
---------
# What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet
though.
Once merged, your PR is going to appear in the release notes with the
title you set, so make sure it's a great title that fully reflects the
extent of your awesome contribution.
Then, please replace this with a description of the change and which
issue is fixed (if applicable). Please also include relevant motivation
and context. List any dependencies (if any) that are required for this
change.
Once you're done, someone will review your PR shortly (see the section
"Who can review?" below to tag some potential reviewers). They may
suggest changes to make the code even better. If no one reviewed your PR
after a week has passed, don't hesitate to post a new comment
@-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Did you read the [contributor
guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the
[forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes?
Here are the
[documentation
guidelines](https://github.com/huggingface/transformers/tree/main/docs),
and
[here are tips on formatting
docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have
passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the
right person to tag with @
@OlivierDehaene OR @Narsil
-->
---------
Co-authored-by: Abhinav M Kulkarni <abhinavkulkarni@gmail.com>
Co-authored-by: Abhinav Kulkarni <abhinav@concentric.ai>
2023-09-25 07:31:27 -06:00
|
|
|
# Copy build artifacts from awq kernels builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=awq-kernels-builder /usr/src/llm-awq/awq/kernels/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-09-29 03:19:06 -06:00
|
|
|
# Copy build artifacts from eetq kernels builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=eetq-kernels-builder /usr/src/eetq/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-04-16 16:26:47 -06:00
|
|
|
|
2023-06-30 11:09:59 -06:00
|
|
|
# Copy builds artifacts from vllm builder
|
2023-11-23 05:38:50 -07:00
|
|
|
COPY --from=vllm-builder /usr/src/vllm/build/lib.linux-x86_64-cpython-310 /opt/conda/lib/python3.10/site-packages
|
2023-06-30 11:09:59 -06:00
|
|
|
|
2024-02-08 02:19:45 -07:00
|
|
|
# Copy build artifacts from mamba builder
|
|
|
|
COPY --from=mamba-builder /usr/src/mamba/build/lib.linux-x86_64-cpython-310/ /opt/conda/lib/python3.10/site-packages
|
|
|
|
COPY --from=mamba-builder /usr/src/causal-conv1d/build/lib.linux-x86_64-cpython-310/ /opt/conda/lib/python3.10/site-packages
|
|
|
|
|
2024-04-10 09:20:25 -06:00
|
|
|
# Install vllm/flash-attention dependencies
|
2023-06-08 06:51:52 -06:00
|
|
|
RUN pip install einops --no-cache-dir
|
2023-04-09 11:59:16 -06:00
|
|
|
|
2022-10-14 07:56:21 -06:00
|
|
|
# Install server
|
2022-10-22 12:00:15 -06:00
|
|
|
COPY proto proto
|
2022-10-14 07:56:21 -06:00
|
|
|
COPY server server
|
2023-04-16 16:26:47 -06:00
|
|
|
COPY server/Makefile server/Makefile
|
2022-10-14 07:56:21 -06:00
|
|
|
RUN cd server && \
|
2022-10-22 12:00:15 -06:00
|
|
|
make gen-server && \
|
2023-11-27 06:08:12 -07:00
|
|
|
pip install -r requirements_cuda.txt && \
|
2024-02-16 09:50:57 -07:00
|
|
|
pip install ".[bnb, accelerate, quantize, peft, outlines]" --no-cache-dir
|
2022-10-14 07:56:21 -06:00
|
|
|
|
2023-05-09 05:19:31 -06:00
|
|
|
# Install benchmarker
|
|
|
|
COPY --from=builder /usr/src/target/release/text-generation-benchmark /usr/local/bin/text-generation-benchmark
|
2022-10-14 07:56:21 -06:00
|
|
|
# Install router
|
2023-03-03 07:07:27 -07:00
|
|
|
COPY --from=builder /usr/src/target/release/text-generation-router /usr/local/bin/text-generation-router
|
2022-10-22 12:00:15 -06:00
|
|
|
# Install launcher
|
2023-03-03 07:07:27 -07:00
|
|
|
COPY --from=builder /usr/src/target/release/text-generation-launcher /usr/local/bin/text-generation-launcher
|
2022-10-14 07:56:21 -06:00
|
|
|
|
feat(server): Add inference support for GPTQ (llama + falcon tested) + Quantization script (#438)
Let's start discussing implementation.
- Need to expose the quantization scripts (either included here or add
doc on how to use https://github.com/qwopqwop200/GPTQ-for-LLaMa)
- Make sure GPTQ works for multiple models (priority to Falcon).
Currently it means that every place we use `get_{tensor|sharded}` to
check for quantization.
My idea is to reintegrate as much as possible into `utils/layer.py` by
expanding `load_multi` to be a bit more generic.
This might require some thinking, but ultimately the
`qweight,qzeros,scales,g_idx` should be in a single place, and
independant of bias presence.
# What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet
though.
Once merged, your PR is going to appear in the release notes with the
title you set, so make sure it's a great title that fully reflects the
extent of your awesome contribution.
Then, please replace this with a description of the change and which
issue is fixed (if applicable). Please also include relevant motivation
and context. List any dependencies (if any) that are required for this
change.
Once you're done, someone will review your PR shortly (see the section
"Who can review?" below to tag some potential reviewers). They may
suggest changes to make the code even better. If no one reviewed your PR
after a week has passed, don't hesitate to post a new comment
@-mentioning the same persons---sometimes notifications get lost.
-->
<!-- Remove if not applicable -->
Fixes # (issue)
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Did you read the [contributor
guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests),
Pull Request section?
- [ ] Was this discussed/approved via a Github issue or the
[forum](https://discuss.huggingface.co/)? Please add a link
to it if that's the case.
- [ ] Did you make sure to update the documentation with your changes?
Here are the
[documentation
guidelines](https://github.com/huggingface/transformers/tree/main/docs),
and
[here are tips on formatting
docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation).
- [ ] Did you write any new necessary tests?
## Who can review?
Anyone in the community is free to review the PR once the tests have
passed. Feel free to tag
members/contributors who may be interested in your PR.
<!-- Your PR will be replied to more quickly if you can figure out the
right person to tag with @
@OlivierDehaene OR @Narsil
-->
---------
Co-authored-by: Ubuntu <ubuntu@ip-172-31-41-161.ec2.internal>
Co-authored-by: OlivierDehaene <olivier@huggingface.co>
2023-06-26 04:27:01 -06:00
|
|
|
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
|
|
|
|
build-essential \
|
|
|
|
g++ \
|
|
|
|
&& rm -rf /var/lib/apt/lists/*
|
|
|
|
|
2023-11-27 06:08:12 -07:00
|
|
|
# AWS Sagemaker compatible image
|
2023-03-29 13:38:30 -06:00
|
|
|
FROM base as sagemaker
|
|
|
|
|
|
|
|
COPY sagemaker-entrypoint.sh entrypoint.sh
|
|
|
|
RUN chmod +x entrypoint.sh
|
|
|
|
|
|
|
|
ENTRYPOINT ["./entrypoint.sh"]
|
|
|
|
|
2023-04-14 02:12:21 -06:00
|
|
|
# Final image
|
2023-03-29 13:38:30 -06:00
|
|
|
FROM base
|
|
|
|
|
2023-02-03 04:43:37 -07:00
|
|
|
ENTRYPOINT ["text-generation-launcher"]
|
2023-06-08 06:51:52 -06:00
|
|
|
CMD ["--json-output"]
|