Add codegen and docsum CI (#51)
Signed-off-by: Sun, Xuehao <xuehao.sun@intel.com> Co-authored-by: chensuyue <suyue.chen@intel.com>
This commit is contained in:
Sun, Xuehao
43
.github/workflows/CodeGen.yml
vendored
Normal file
43
.github/workflows/CodeGen.yml
vendored
Normal file
@@ -0,0 +1,43 @@
|
||||
name: CodeGen-test
|
||||
|
||||
on:
|
||||
pull_request:
|
||||
branches: [main]
|
||||
types: [opened, reopened, ready_for_review, synchronize] # added `ready_for_review` since draft is skipped
|
||||
paths:
|
||||
- CodeGen/**
|
||||
- .github/workflows/CodeGen.yml
|
||||
workflow_dispatch:
|
||||
|
||||
# If there is a new commit, the previous jobs will be canceled
|
||||
concurrency:
|
||||
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
|
||||
cancel-in-progress: true
|
||||
|
||||
jobs:
|
||||
CodeGen:
|
||||
runs-on: gaudi2
|
||||
strategy:
|
||||
matrix:
|
||||
job_name: ["codegen"]
|
||||
fail-fast: false
|
||||
steps:
|
||||
- name: Clean Up Working Directory
|
||||
run: sudo rm -rf ${{github.workspace}}/*
|
||||
|
||||
- name: Checkout out Repo
|
||||
uses: actions/checkout@v4
|
||||
|
||||
- name: Run Test
|
||||
env:
|
||||
HUGGINGFACEHUB_API_TOKEN: ${{ secrets.HUGGINGFACEHUB_API_TOKEN }}
|
||||
run: |
|
||||
cd ${{ github.workspace }}/CodeGen/tests
|
||||
bash test_${{ matrix.job_name }}_inference.sh
|
||||
|
||||
- name: Publish pipeline artifact
|
||||
if: ${{ !cancelled() }}
|
||||
uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: ${{ matrix.job_name }}
|
||||
path: ${{ github.workspace }}/CodeGen/tests/*.log
|
||||
43
.github/workflows/DocSum.yml
vendored
Normal file
43
.github/workflows/DocSum.yml
vendored
Normal file
@@ -0,0 +1,43 @@
|
||||
name: DocSum-test
|
||||
|
||||
on:
|
||||
pull_request:
|
||||
branches: [main]
|
||||
types: [opened, reopened, ready_for_review, synchronize] # added `ready_for_review` since draft is skipped
|
||||
paths:
|
||||
- DocSum/**
|
||||
- .github/workflows/DocSum.yml
|
||||
workflow_dispatch:
|
||||
|
||||
# If there is a new commit, the previous jobs will be canceled
|
||||
concurrency:
|
||||
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
|
||||
cancel-in-progress: true
|
||||
|
||||
jobs:
|
||||
DocSum:
|
||||
runs-on: gaudi2
|
||||
strategy:
|
||||
matrix:
|
||||
job_name: ["langchain"]
|
||||
fail-fast: false
|
||||
steps:
|
||||
- name: Clean Up Working Directory
|
||||
run: sudo rm -rf ${{github.workspace}}/*
|
||||
|
||||
- name: Checkout out Repo
|
||||
uses: actions/checkout@v4
|
||||
|
||||
- name: Run Test
|
||||
env:
|
||||
HUGGINGFACEHUB_API_TOKEN: ${{ secrets.HUGGINGFACEHUB_API_TOKEN }}
|
||||
run: |
|
||||
cd ${{ github.workspace }}/DocSum/tests
|
||||
bash test_${{ matrix.job_name }}_inference.sh
|
||||
|
||||
- name: Publish pipeline artifact
|
||||
if: ${{ !cancelled() }}
|
||||
uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: ${{ matrix.job_name }}
|
||||
path: ${{ github.workspace }}/DocSum/tests/*.log
|
||||
@@ -103,7 +103,7 @@ function docker_stop() {
|
||||
function main() {
|
||||
test_env_setup
|
||||
rename
|
||||
docker_stop $CHATQNA_CONTAINER_NAME && docker_stop $LANGCHAIN_CONTAINER_NAME && docker_stop $REDIS_CONTAINER_NAME
|
||||
docker_stop $CHATQNA_CONTAINER_NAME && docker_stop $LANGCHAIN_CONTAINER_NAME && docker_stop $REDIS_CONTAINER_NAME && sleep 5s
|
||||
|
||||
launch_tgi_gaudi_service
|
||||
launch_redis_and_langchain_service
|
||||
@@ -111,7 +111,7 @@ function main() {
|
||||
|
||||
run_tests
|
||||
|
||||
docker_stop $CHATQNA_CONTAINER_NAME && docker_stop $LANGCHAIN_CONTAINER_NAME && docker_stop $REDIS_CONTAINER_NAME
|
||||
docker_stop $CHATQNA_CONTAINER_NAME && docker_stop $LANGCHAIN_CONTAINER_NAME && docker_stop $REDIS_CONTAINER_NAME && sleep 5s
|
||||
echo y | docker system prune
|
||||
|
||||
check_response
|
||||
|
||||
@@ -41,9 +41,9 @@ volume=$PWD/data
|
||||
|
||||
# Build the Docker run command based on the number of cards
|
||||
if [ "$num_cards" -eq 1 ]; then
|
||||
docker_cmd="docker run -p $port_number:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name"
|
||||
docker_cmd="docker run -d --name="CodeGen_server" -p $port_number:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name"
|
||||
else
|
||||
docker_cmd="docker run -p $port_number:80 -v $volume:/data --runtime=habana -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name --sharded true --num-shard $num_cards"
|
||||
docker_cmd="docker run -d --name="CodeGen_server" -p $port_number:80 -v $volume:/data --runtime=habana -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name --sharded true --num-shard $num_cards"
|
||||
fi
|
||||
|
||||
# Execute the Docker run command
|
||||
|
||||
113
CodeGen/tests/test_codegen_inference.sh
Normal file
113
CodeGen/tests/test_codegen_inference.sh
Normal file
@@ -0,0 +1,113 @@
|
||||
#!/bin/bash
|
||||
# Copyright (c) 2024 Intel Corporation
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
set -xe
|
||||
|
||||
function test_env_setup() {
|
||||
WORKPATH=$(dirname "$PWD")
|
||||
LOG_PATH="$WORKPATH/tests/codegen.log"
|
||||
|
||||
COPILOT_CONTAINER_NAME="test-copilot"
|
||||
CODEGEN_CONTAINER_NAME="test-CodeGen_server"
|
||||
cd $WORKPATH # go to CodeGen
|
||||
}
|
||||
|
||||
function rename() {
|
||||
# Rename the container names
|
||||
cd ${WORKPATH}
|
||||
sed -i "s/CodeGen_server/${CODEGEN_CONTAINER_NAME}/g" serving/tgi_gaudi/launch_tgi_service.sh
|
||||
}
|
||||
|
||||
function docker_setup() {
|
||||
local card_num=1
|
||||
local port=8902
|
||||
local model_name="m-a-p/OpenCodeInterpreter-DS-6.7B"
|
||||
|
||||
cd ${WORKPATH}
|
||||
|
||||
# Reset the tgi port
|
||||
sed -i "s/8080/$port/g" codegen/codegen-app/server.py
|
||||
|
||||
docker pull ghcr.io/huggingface/tgi-gaudi:1.2.1
|
||||
bash serving/tgi_gaudi/launch_tgi_service.sh $card_num $port $model_name
|
||||
sleep 3m # Waits 3 minutes
|
||||
}
|
||||
|
||||
function launch_copilot_docker() {
|
||||
local port=8903
|
||||
sed -i "s/port=8000/port=$port/g" codegen/codegen-app/server.py
|
||||
|
||||
cd $WORKPATH/codegen
|
||||
bash ./build_docker.sh
|
||||
|
||||
cd $WORKPATH
|
||||
docker run -dit --name=$COPILOT_CONTAINER_NAME \
|
||||
--net=host --ipc=host \
|
||||
-v /var/run/docker.sock:/var/run/docker.sock intel/gen-ai-examples:copilot /bin/bash
|
||||
}
|
||||
|
||||
function launch_server() {
|
||||
cd $WORKPATH
|
||||
|
||||
# Start the Backend Service
|
||||
docker exec $COPILOT_CONTAINER_NAME \
|
||||
bash -c "export HUGGINGFACEHUB_API_TOKEN=$HUGGINGFACEHUB_API_TOKEN;nohup python server.py &"
|
||||
sleep 1m
|
||||
}
|
||||
|
||||
function run_tests() {
|
||||
cd $WORKPATH
|
||||
local port=8903
|
||||
}
|
||||
|
||||
function check_response() {
|
||||
cd $WORKPATH
|
||||
echo "Checking response"
|
||||
local status=false
|
||||
if [[ $(grep -c "\$51.2 billion" $LOG_PATH) != 0 ]]; then
|
||||
status=true
|
||||
fi
|
||||
|
||||
if [ $status == false ]; then
|
||||
echo "Response check failed"
|
||||
exit 1
|
||||
else
|
||||
echo "Response check succeed"
|
||||
fi
|
||||
}
|
||||
|
||||
function docker_stop() {
|
||||
local container_name=$1
|
||||
cid=$(docker ps -aq --filter "name=$container_name")
|
||||
if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
|
||||
}
|
||||
|
||||
function main() {
|
||||
test_env_setup
|
||||
rename
|
||||
docker_stop $CODEGEN_CONTAINER_NAME && docker_stop $COPILOT_CONTAINER_NAME && sleep 5s
|
||||
|
||||
docker_setup
|
||||
launch_copilot_docker
|
||||
launch_server
|
||||
|
||||
run_tests
|
||||
|
||||
docker_stop $CODEGEN_CONTAINER_NAME && docker_stop $COPILOT_CONTAINER_NAME && sleep 5s
|
||||
echo y | docker system prune
|
||||
|
||||
check_response
|
||||
}
|
||||
|
||||
main
|
||||
@@ -14,4 +14,4 @@
|
||||
# limitations under the License.
|
||||
|
||||
|
||||
docker build . -t document-summarize:latest --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
|
||||
docker build . -t intel/gen-ai-examples:document-summarize --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
|
||||
|
||||
@@ -40,9 +40,9 @@ volume=$PWD/data
|
||||
|
||||
# Build the Docker run command based on the number of cards
|
||||
if [ "$num_cards" -eq 1 ]; then
|
||||
docker_cmd="docker run -p $port_number:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name"
|
||||
docker_cmd="docker run -d --name="DocSum_server" -p $port_number:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name"
|
||||
else
|
||||
docker_cmd="docker run -p $port_number:80 -v $volume:/data --runtime=habana -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name --sharded true --num-shard $num_cards"
|
||||
docker_cmd="docker run -d --name="DocSum_server" -p $port_number:80 -v $volume:/data --runtime=habana -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model_name --sharded true --num-shard $num_cards"
|
||||
fi
|
||||
|
||||
# Execute the Docker run command
|
||||
|
||||
121
DocSum/tests/test_langchain_inference.sh
Normal file
121
DocSum/tests/test_langchain_inference.sh
Normal file
@@ -0,0 +1,121 @@
|
||||
#!/bin/bash
|
||||
# Copyright (c) 2024 Intel Corporation
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
set -xe
|
||||
|
||||
function test_env_setup() {
|
||||
WORKPATH=$(dirname "$PWD")
|
||||
LOG_PATH="$WORKPATH/tests/langchain.log"
|
||||
|
||||
DOCUMENT_SUMMARY_CONTAINER_NAME="test-document-summary"
|
||||
DOCSUM_CONTAINER_NAME="test-DocSum_server"
|
||||
cd $WORKPATH # go to DocSum
|
||||
}
|
||||
|
||||
function rename() {
|
||||
# Rename the container names
|
||||
cd ${WORKPATH}
|
||||
sed -i "s/DocSum_server/${DOCSUM_CONTAINER_NAME}/g" serving/tgi_gaudi/launch_tgi_service.sh
|
||||
}
|
||||
|
||||
function docker_setup() {
|
||||
local card_num=1
|
||||
local port=8900
|
||||
local model_name="Intel/neural-chat-7b-v3-3"
|
||||
|
||||
cd ${WORKPATH}
|
||||
|
||||
# Reset the tgi port
|
||||
sed -i "s/8080/$port/g" langchain/docker/summarize-app/app/server.py
|
||||
sed -i "s/8080/$port/g" langchain/docker/summarize-app/Dockerfile
|
||||
|
||||
docker pull ghcr.io/huggingface/tgi-gaudi:1.2.1
|
||||
bash serving/tgi_gaudi/launch_tgi_service.sh $card_num $port $model_name
|
||||
sleep 3m # Waits 3 minutes
|
||||
}
|
||||
|
||||
function launch_document_summary_docker() {
|
||||
local port=8901
|
||||
sed -i "s/port=8000/port=$port/g" langchain/docker/summarize-app/app/server.py
|
||||
|
||||
cd $WORKPATH/langchain/docker/
|
||||
bash ./build_docker.sh
|
||||
|
||||
cd $WORKPATH
|
||||
docker run -dit --net=host --ipc=host \
|
||||
--name=$DOCUMENT_SUMMARY_CONTAINER_NAME \
|
||||
-v /var/run/docker.sock:/var/run/docker.sock intel/gen-ai-examples:document-summarize /bin/bash
|
||||
}
|
||||
|
||||
function launch_server() {
|
||||
cd $WORKPATH
|
||||
|
||||
# Start the Backend Service
|
||||
docker exec $DOCUMENT_SUMMARY_CONTAINER_NAME \
|
||||
bash -c "export HUGGINGFACEHUB_API_TOKEN=$HUGGINGFACEHUB_API_TOKEN;nohup python app/server.py &"
|
||||
sleep 1m
|
||||
}
|
||||
|
||||
function run_tests() {
|
||||
cd $WORKPATH
|
||||
local port=8901
|
||||
|
||||
status_code=$(curl -s -o /dev/null -w "%{http_code}" http://127.0.0.1:$port/v1/text_summarize \
|
||||
-X POST \
|
||||
-H 'Content-Type: application/json' \
|
||||
-d '{"text":"Text Embeddings Inference (TEI) is a toolkit for deploying and serving open source text embeddings and sequence classification models. TEI enables high-performance extraction for the most popular models, including FlagEmbedding, Ember, GTE and E5."}') || true
|
||||
|
||||
sleep 5s
|
||||
}
|
||||
|
||||
function check_response() {
|
||||
cd $WORKPATH
|
||||
echo "Checking response"
|
||||
local status=false
|
||||
if [ "$status_code" -eq 200 ]; then
|
||||
status=true
|
||||
fi
|
||||
|
||||
if [ $status == false ]; then
|
||||
echo "Response check failed"
|
||||
exit 1
|
||||
else
|
||||
echo "Response check succeed"
|
||||
fi
|
||||
}
|
||||
|
||||
function docker_stop() {
|
||||
local container_name=$1
|
||||
cid=$(docker ps -aq --filter "name=$container_name")
|
||||
if [[ ! -z "$cid" ]]; then docker stop $cid && docker rm $cid; fi
|
||||
}
|
||||
|
||||
function main() {
|
||||
test_env_setup
|
||||
rename
|
||||
docker_stop $DOCSUM_CONTAINER_NAME && docker_stop $DOCUMENT_SUMMARY_CONTAINER_NAME && sleep 5s
|
||||
|
||||
docker_setup
|
||||
launch_document_summary_docker
|
||||
launch_server
|
||||
|
||||
run_tests
|
||||
|
||||
docker_stop $DOCSUM_CONTAINER_NAME && docker_stop $DOCUMENT_SUMMARY_CONTAINER_NAME && sleep 5s
|
||||
echo y | docker system prune
|
||||
|
||||
check_response
|
||||
}
|
||||
|
||||
main
|
||||
Reference in New Issue
Block a user