feat: migrate python backends from conda to uv (#2215)

* feat: migrate diffusers backend from conda to uv

  - replace conda with UV for diffusers install (prototype for all
    extras backends)
  - add ability to build docker with one/some/all extras backends
    instead of all or nothing

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate autogtpq bark coqui from conda to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: convert exllama over to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate exllama2 to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate mamba to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate parler to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate petals to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: fix tests

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate rerankers to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate sentencetransformers to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: install uv for tests-linux

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: make sure file exists before installing on intel images

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate transformers backend to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate transformers-musicgen to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate vall-e-x to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: migrate vllm to uv

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: add uv install to the rest of test-extra.yml

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: adjust file perms on all install/run/test scripts

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: add missing acclerate dependencies

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: add some more missing dependencies to python backends

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: parler tests venv py dir fix

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: correct filename for transformers-musicgen tests

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: adjust the pwd for valle tests

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: cleanup and optimization work for uv migration

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: add setuptools to requirements-install for mamba

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: more size optimization work

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* feat: make installs and tests more consistent, cleanup some deps

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: cleanup

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: mamba backend is cublas only

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

* fix: uncomment lines in makefile

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>

---------

Signed-off-by: Chris Jowett <421501+cryptk@users.noreply.github.com>
This commit is contained in:
cryptk 2024-05-10 08:08:08 -05:00 committed by GitHub
parent e6768097f4
commit 28a421cb1d
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
101 changed files with 988 additions and 740 deletions

View file

@ -1,6 +1,6 @@
.PHONY: vllm
vllm: protogen
$(MAKE) -C ../common-env/transformers
bash install.sh
.PHONY: run
run: protogen
@ -22,4 +22,8 @@ protogen-clean:
$(RM) backend_pb2_grpc.py backend_pb2.py
backend_pb2_grpc.py backend_pb2.py:
python3 -m grpc_tools.protoc -I../.. --python_out=. --grpc_python_out=. backend.proto
python3 -m grpc_tools.protoc -I../.. --python_out=. --grpc_python_out=. backend.proto
.PHONY: clean
clean: protogen-clean
rm -rf venv

34
backend/python/vllm/install.sh Executable file
View file

@ -0,0 +1,34 @@
#!/bin/bash
set -ex
BUILD_ISOLATION_FLAG=""
MY_DIR="$(dirname -- "${BASH_SOURCE[0]}")"
uv venv ${MY_DIR}/venv
source ${MY_DIR}/venv/bin/activate
if [ -f "requirements-install.txt" ]; then
# If we have a requirements-install.txt, it means that a package does not properly declare it's build time
# dependencies per PEP-517, so we have to set up the proper build environment ourselves, and then install
# the package without build isolation
BUILD_ISOLATION_FLAG="--no-build-isolation"
uv pip install --requirement ${MY_DIR}/requirements-install.txt
fi
uv pip install ${BUILD_ISOLATION_FLAG} --requirement ${MY_DIR}/requirements.txt
if [ -f "requirements-${BUILD_TYPE}.txt" ]; then
uv pip install ${BUILD_ISOLATION_FLAG} --requirement ${MY_DIR}/requirements-${BUILD_TYPE}.txt
fi
if [ -d "/opt/intel" ]; then
# Intel GPU: If the directory exists, we assume we are using the Intel image
# https://github.com/intel/intel-extension-for-pytorch/issues/538
if [ -f "requirements-intel.txt" ]; then
uv pip install ${BUILD_ISOLATION_FLAG} --index-url https://pytorch-extension.intel.com/release-whl/stable/xpu/us/ --requirement ${MY_DIR}/requirements-intel.txt
fi
fi
if [ "$PIP_CACHE_PURGE" = true ] ; then
pip cache purge
fi

View file

@ -0,0 +1 @@
flash-attn

View file

@ -0,0 +1,6 @@
# mabma does not specify it's build dependencies per PEP517, so we need to disable build isolation
# this also means that we need to install the basic build dependencies into the venv ourselves
# https://github.com/Dao-AILab/causal-conv1d/issues/24
packaging
setuptools
wheel

View file

@ -0,0 +1,7 @@
accelerate
vllm
grpcio==1.63.0
protobuf
certifi
transformers
setuptools

View file

@ -1,14 +1,10 @@
#!/bin/bash
##
## A bash script wrapper that runs the diffusers server with conda
## A bash script wrapper that runs the GRPC backend
export PATH=$PATH:/opt/conda/bin
MY_DIR="$(dirname -- "${BASH_SOURCE[0]}")"
# Activate conda environment
source activate transformers
source $MY_DIR/venv/bin/activate
# get the directory where the bash script is located
DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
python $DIR/backend_vllm.py $@
python $MY_DIR/backend_vllm.py $@

17
backend/python/vllm/test.sh Normal file → Executable file
View file

@ -1,11 +1,16 @@
#!/bin/bash
##
## A bash script wrapper that runs the transformers server with conda
## A bash script wrapper that runs python unittests
# Activate conda environment
source activate transformers
MY_DIR="$(dirname -- "${BASH_SOURCE[0]}")"
# get the directory where the bash script is located
DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
source $MY_DIR/venv/bin/activate
python -m unittest $DIR/test_backend_vllm.py
if [ -f "${MY_DIR}/test.py" ]; then
pushd ${MY_DIR}
python -m unittest test.py
popd
else
echo "ERROR: No tests defined for backend!"
exit 1
fi