Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove enforce-eager to enable HPU graphs for better vLLM perf #954

Merged
merged 4 commits into from
Dec 10, 2024
Merged

Remove enforce-eager to enable HPU graphs for better vLLM perf #954

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
6ab8cc7
remove enforce-eager to enable HPU graphs
wangkl2 Nov 28, 2024
9f4ee4e
Merge branch 'main' into update-vllm-gaudi
wangkl2 Dec 9, 2024
b50e834
Merge branch 'opea-project:main' into update-vllm-gaudi
wangkl2 Dec 10, 2024
3828ff8
Increase the llm max timeout in ci for fully warmup
wangkl2 Dec 10, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion comps/llms/faq-generation/vllm/langchain/docker_compose_llm.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ services:
cap_add:
- SYS_NICE
ipc: host
command: --enforce-eager --model $LLM_MODEL_ID --tensor-parallel-size 1 --host 0.0.0.0 --port 80
command: --model $LLM_MODEL_ID --tensor-parallel-size 1 --host 0.0.0.0 --port 80
llm:
image: opea/llm-faqgen-vllm:latest
container_name: llm-faqgen-server
Expand Down
2 changes: 1 addition & 1 deletion comps/llms/summarization/vllm/langchain/docker_compose_llm.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ services:
cap_add:
- SYS_NICE
ipc: host
command: --enforce-eager --model $LLM_MODEL_ID --tensor-parallel-size 1 --host 0.0.0.0 --port 80
command: --model $LLM_MODEL_ID --tensor-parallel-size 1 --host 0.0.0.0 --port 80
llm:
image: opea/llm-docsum-vllm:latest
container_name: llm-docsum-vllm-server
Expand Down
2 changes: 1 addition & 1 deletion comps/llms/text-generation/vllm/langchain/dependency/launch_vllm_service.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ volume=$PWD/data

# Build the Docker run command based on hardware mode
if [ "$hw_mode" = "hpu" ]; then
docker run -d --rm --runtime=habana --name="vllm-service" -p $port_number:80 -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HF_TOKEN} opea/vllm-gaudi:latest --enforce-eager --model $model_name --tensor-parallel-size $parallel_number --host 0.0.0.0 --port 80 --block-size $block_size --max-num-seqs $max_num_seqs --max-seq_len-to-capture $max_seq_len_to_capture
docker run -d --rm --runtime=habana --name="vllm-service" -p $port_number:80 -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HF_TOKEN} opea/vllm-gaudi:latest --model $model_name --tensor-parallel-size $parallel_number --host 0.0.0.0 --port 80 --block-size $block_size --max-num-seqs $max_num_seqs --max-seq_len-to-capture $max_seq_len_to_capture
else
docker run -d --rm --name="vllm-service" -p $port_number:80 --network=host -v $volume:/data -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HF_TOKEN} -e VLLM_CPU_KVCACHE_SPACE=40 opea/vllm-cpu:latest --model $model_name --host 0.0.0.0 --port 80
fi
2 changes: 1 addition & 1 deletion comps/llms/text-generation/vllm/langchain/docker_compose_llm.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ services:
cap_add:
- SYS_NICE
ipc: host
command: --enforce-eager --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80
command: --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80
llm:
image: opea/llm-vllm:latest
container_name: llm-vllm-gaudi-server
Expand Down
2 changes: 1 addition & 1 deletion comps/llms/text-generation/vllm/llama_index/dependency/launch_vllm_service.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ volume=$PWD/data

# Build the Docker run command based on hardware mode
if [ "$hw_mode" = "hpu" ]; then
docker run -d --rm --runtime=habana --name="vllm-service" -p $port_number:80 -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN} opea/vllm-gaudi:latest --enforce-eager --model $model_name --tensor-parallel-size $parallel_number --host 0.0.0.0 --port 80 --block-size $block_size --max-num-seqs $max_num_seqs --max-seq_len-to-capture $max_seq_len_to_capture
docker run -d --rm --runtime=habana --name="vllm-service" -p $port_number:80 -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN} opea/vllm-gaudi:latest --model $model_name --tensor-parallel-size $parallel_number --host 0.0.0.0 --port 80 --block-size $block_size --max-num-seqs $max_num_seqs --max-seq_len-to-capture $max_seq_len_to_capture
else
docker run -d --rm --name="vllm-service" -p $port_number:80 --network=host -v $volume:/data -e HTTPS_PROXY=$https_proxy -e HTTP_PROXY=$https_proxy -e HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN} -e VLLM_CPU_KVCACHE_SPACE=40 opea/vllm-cpu:latest --model $model_name --host 0.0.0.0 --port 80
fi
2 changes: 1 addition & 1 deletion comps/llms/text-generation/vllm/llama_index/docker_compose_llm.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ services:
cap_add:
- SYS_NICE
ipc: host
command: --enforce-eager --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80
command: --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80
llm:
image: opea/llm-vllm-llamaindex:latest
container_name: llm-vllm-gaudi-server
Expand Down
4 changes: 2 additions & 2 deletions tests/llms/test_llms_text-generation_vllm_langchain_on_intel_hpu.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,7 @@ function start_service() {
--ipc=host \
-e HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN} \
opea/vllm-gaudi:comps \
--enforce-eager --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80 --block-size 128 --max-num-seqs 256 --max-seq_len-to-capture 2048
--model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80 --block-size 128 --max-num-seqs 256 --max-seq_len-to-capture 2048

export vLLM_ENDPOINT="http://${ip_address}:${port_number}"
docker run -d --rm \
Expand All @@ -62,7 +62,7 @@ function start_service() {

# check whether vllm ray is fully ready
n=0
until [[ "$n" -ge 120 ]] || [[ $ready == true ]]; do
until [[ "$n" -ge 160 ]] || [[ $ready == true ]]; do
docker logs test-comps-vllm-service > ${WORKPATH}/tests/test-comps-vllm-service.log
n=$((n+1))
if grep -q throughput ${WORKPATH}/tests/test-comps-vllm-service.log; then
Expand Down
4 changes: 2 additions & 2 deletions tests/llms/test_llms_text-generation_vllm_llamaindex_on_intel_hpu.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,7 @@ function start_service() {
--ipc=host \
-e HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN} \
opea/vllm-gaudi:comps \
--enforce-eager --model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80 --block-size 128 --max-num-seqs 256 --max-seq_len-to-capture 2048
--model $LLM_MODEL --tensor-parallel-size 1 --host 0.0.0.0 --port 80 --block-size 128 --max-num-seqs 256 --max-seq_len-to-capture 2048

export vLLM_ENDPOINT="http://${ip_address}:${port_number}"
docker run -d --rm \
Expand All @@ -62,7 +62,7 @@ function start_service() {

# check whether vllm ray is fully ready
n=0
until [[ "$n" -ge 120 ]] || [[ $ready == true ]]; do
until [[ "$n" -ge 160 ]] || [[ $ready == true ]]; do
docker logs test-comps-vllm-service > ${WORKPATH}/tests/test-comps-vllm-service.log
n=$((n+1))
if grep -q throughput ${WORKPATH}/tests/test-comps-vllm-service.log; then
Expand Down
Loading