[Bug] `llm-docsum` test fail on vllm backend

### Priority

P1-Stopper

### OS type

N/A

### Hardware type

N/A

### Installation method

- [ ] Pull docker images from hub.docker.com
- [ ] Build docker images from source
- [ ] Other
- [x] N/A

### Deploy method

- [ ] Docker
- [ ] Docker Compose
- [ ] Kubernetes Helm Charts
- [ ] Other
- [x] N/A

### Running nodes

N/A

### What's the version?

main

### Description

The test hangs and eventually times out after the following part of the test.


### Reproduce steps

N/A

### Raw log

```shell
2025-11-19T08:04:28.4244544Z + docker compose -f compose_doc-summarization.yaml up docsum-vllm -d
2025-11-19T08:04:28.5069115Z time="2025-11-19T08:04:28Z" level=warning msg="The \"offline_no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5072210Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5075137Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5077481Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5080353Z time="2025-11-19T08:04:28Z" level=warning msg="The \"offline_no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5082573Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5084911Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5087051Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5089176Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5090767Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5091458Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5181247Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5182709Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5184245Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5185657Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5186933Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5188151Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5189485Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5190763Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5192064Z time="2025-11-19T08:04:28Z" level=warning msg="The \"VLLM_LLM_MODEL_ID\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5193430Z time="2025-11-19T08:04:28Z" level=warning msg="The \"TENSOR_PARALLEL_SIZE\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5194990Z time="2025-11-19T08:04:28Z" level=warning msg="The \"no_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5196305Z time="2025-11-19T08:04:28Z" level=warning msg="The \"http_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5197652Z time="2025-11-19T08:04:28Z" level=warning msg="The \"https_proxy\" variable is not set. Defaulting to a blank string."
2025-11-19T08:04:28.5496785Z  Network docker_compose_default  Creating
2025-11-19T08:04:28.6089152Z  Network docker_compose_default  Created
2025-11-19T08:04:28.6099567Z  Container vllm-server  Creating
2025-11-19T08:04:28.6669243Z  Container vllm-server  Created
2025-11-19T08:04:28.6675883Z  Container docsum-vllm  Creating
2025-11-19T08:04:28.6899651Z  Container docsum-vllm  Created
2025-11-19T08:04:28.6913203Z  Container vllm-server  Starting
2025-11-19T08:04:28.8764352Z  Container vllm-server  Started
2025-11-19T08:04:28.8765514Z  Container vllm-server  Waiting
2025-11-19T08:07:00.3790563Z  Container vllm-server  Healthy
2025-11-19T08:07:00.3791690Z  Container docsum-vllm  Starting
2025-11-19T08:07:00.5791577Z  Container docsum-vllm  Started
2025-11-19T08:07:00.5868904Z + sleep 30s
2025-11-19T08:07:30.5905678Z + validate_microservices
2025-11-19T08:07:30.5906918Z Validate vllm...
2025-11-19T08:07:30.5908177Z + URL=http://192.168.122.213:10507/v1/docsum
2025-11-19T08:07:30.5909857Z + echo 'Validate vllm...'
2025-11-19T08:07:30.5914219Z + validate_services http://192.168.122.213:12107/v1/completions text vllm-server vllm-server '{"model": "meta-llama/Meta-Llama-3-8B-Instruct", "prompt": "What is Deep Learning?", "max_tokens": 32, "temperature": 0}'
2025-11-19T08:07:30.5915133Z + local URL=http://192.168.122.213:12107/v1/completions
2025-11-19T08:07:30.5915437Z + local EXPECTED_RESULT=text
2025-11-19T08:07:30.5915725Z + local SERVICE_NAME=vllm-server
2025-11-19T08:07:30.5916011Z + local DOCKER_NAME=vllm-server
2025-11-19T08:07:30.5916550Z + local 'INPUT_DATA={"model": "meta-llama/Meta-Llama-3-8B-Instruct", "prompt": "What is Deep Learning?", "max_tokens": 32, "temperature": 0}'
2025-11-19T08:07:30.5921986Z ++ curl -s -o /dev/null -w '%{http_code}' -X POST -d '{"model": "meta-llama/Meta-Llama-3-8B-Instruct", "prompt": "What is Deep Learning?", "max_tokens": 32, "temperature": 0}' -H 'Content-Type: application/json' http://192.168.122.213:12107/v1/completions
2025-11-19T08:07:35.1660227Z + local HTTP_STATUS=200
2025-11-19T08:07:35.1661259Z + echo ===========================================
2025-11-19T08:07:35.1661760Z + '[' 200 -eq 200 ']'
2025-11-19T08:07:35.1662141Z + echo '[ vllm-server ] HTTP status is 200. Checking content...'
2025-11-19T08:07:35.1662658Z ===========================================
2025-11-19T08:07:35.1663052Z [ vllm-server ] HTTP status is 200. Checking content...
2025-11-19T08:07:35.1666317Z ++ curl -s -X POST -d '{"model": "meta-llama/Meta-Llama-3-8B-Instruct", "prompt": "What is Deep Learning?", "max_tokens": 32, "temperature": 0}' -H 'Content-Type: application/json' http://192.168.122.213:12107/v1/completions
2025-11-19T08:07:35.1668016Z ++ tee /home/sdp/opea-actions-runner/_work/GenAIComps/GenAIComps/tests/vllm-server.log
2025-11-19T08:07:39.7705476Z + local 'CONTENT={"id":"cmpl-144f59bf97e94cdda5bdcba7f47f1c22","object":"text_completion","created":1763539655,"model":"meta-llama/Meta-Llama-3-8B-Instruct","choices":[{"index":0,"text":" A Beginner'\''s Guide\nDeep learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. These neural networks are designed","logprobs":null,"finish_reason":"length","stop_reason":null,"prompt_logprobs":null}],"service_tier":null,"system_fingerprint":null,"usage":{"prompt_tokens":6,"total_tokens":38,"completion_tokens":32,"prompt_tokens_details":null},"kv_transfer_params":null}'
2025-11-19T08:07:39.7709491Z + echo '{"id":"cmpl-144f59bf97e94cdda5bdcba7f47f1c22","object":"text_completion","created":1763539655,"model":"meta-llama/Meta-Llama-3-8B-Instruct","choices":[{"index":0,"text":"' A 'Beginner'\''s' 'Guide\nDeep' learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. These neural networks are 'designed","logprobs":null,"finish_reason":"length","stop_reason":null,"prompt_logprobs":null}],"service_tier":null,"system_fingerprint":null,"usage":{"prompt_tokens":6,"total_tokens":38,"completion_tokens":32,"prompt_tokens_details":null},"kv_transfer_params":null}'
2025-11-19T08:07:39.7713058Z {"id":"cmpl-144f59bf97e94cdda5bdcba7f47f1c22","object":"text_completion","created":1763539655,"model":"meta-llama/Meta-Llama-3-8B-Instruct","choices":[{"index":0,"text":" A Beginner's Guide\nDeep learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. These neural networks are designed","logprobs":null,"finish_reason":"length","stop_reason":null,"prompt_logprobs":null}],"service_tier":null,"system_fingerprint":null,"usage":{"prompt_tokens":6,"total_tokens":38,"completion_tokens":32,"prompt_tokens_details":null},"kv_transfer_params":null}
2025-11-19T08:07:39.7717124Z + echo '{"id":"cmpl-144f59bf97e94cdda5bdcba7f47f1c22","object":"text_completion","created":1763539655,"model":"meta-llama/Meta-Llama-3-8B-Instruct","choices":[{"index":0,"text":" A Beginner'\''s Guide\nDeep learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. These neural networks are designed","logprobs":null,"finish_reason":"length","stop_reason":null,"prompt_logprobs":null}],"service_tier":null,"system_fingerprint":null,"usage":{"prompt_tokens":6,"total_tokens":38,"completion_tokens":32,"prompt_tokens_details":null},"kv_transfer_params":null}'
2025-11-19T08:07:39.7718993Z + grep -q text
2025-11-19T08:07:39.7723310Z + echo '[ vllm-server ] Content is as expected.'
2025-11-19T08:07:39.7723610Z [ vllm-server ] Content is as expected.
2025-11-19T08:07:39.7724069Z + sleep 1s
2025-11-19T08:07:40.7745184Z + echo 'Validate stream=True...'
2025-11-19T08:07:40.7746795Z Validate stream=True...
2025-11-19T08:07:40.7752269Z + validate_services http://192.168.122.213:10507/v1/docsum text docsum-vllm docsum-vllm '{"messages": "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.", "max_tokens":32, "language":"en"}'
2025-11-19T08:07:40.7758646Z + local URL=http://192.168.122.213:10507/v1/docsum
2025-11-19T08:07:40.7760137Z + local EXPECTED_RESULT=text
2025-11-19T08:07:40.7761357Z + local SERVICE_NAME=docsum-vllm
2025-11-19T08:07:40.7762316Z + local DOCKER_NAME=docsum-vllm
2025-11-19T08:07:40.7765189Z + local 'INPUT_DATA={"messages": "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.", "max_tokens":32, "language":"en"}'
2025-11-19T08:07:40.7767269Z ++ curl -s -o /dev/null -w '%{http_code}' -X POST -d '{"messages": "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.", "max_tokens":32, "language":"en"}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:07:50.3855838Z + local HTTP_STATUS=200
2025-11-19T08:07:50.3857372Z ===========================================
2025-11-19T08:07:50.3858913Z + echo ===========================================
2025-11-19T08:07:50.3860223Z + '[' 200 -eq 200 ']'
2025-11-19T08:07:50.3861494Z + echo '[ docsum-vllm ] HTTP status is 200. Checking content...'
2025-11-19T08:07:50.3863080Z [ docsum-vllm ] HTTP status is 200. Checking content...
2025-11-19T08:07:50.3869488Z ++ curl -s -X POST -d '{"messages": "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.", "max_tokens":32, "language":"en"}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:07:50.3871542Z ++ tee /home/sdp/opea-actions-runner/_work/GenAIComps/GenAIComps/tests/docsum-vllm.log
2025-11-19T08:08:00.0622349Z + local 'CONTENT={"id":"d2b5f1c4f0e0a79d636e0f07c9a2e3ee","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:00.0630303Z {"id":"d2b5f1c4f0e0a79d636e0f07c9a2e3ee","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}
2025-11-19T08:08:00.0638077Z + echo '{"id":"d2b5f1c4f0e0a79d636e0f07c9a2e3ee","text":"' '\nTEI' is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, 'Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:00.0641304Z + echo '{"id":"d2b5f1c4f0e0a79d636e0f07c9a2e3ee","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:00.0643101Z + grep -q text
2025-11-19T08:08:00.0643439Z + echo '[ docsum-vllm ] Content is as expected.'
2025-11-19T08:08:00.0643710Z + sleep 1s
2025-11-19T08:08:00.0644106Z [ docsum-vllm ] Content is as expected.
2025-11-19T08:08:01.0674998Z + echo 'Validate stream=False...'
2025-11-19T08:08:01.0676602Z Validate stream=False...
2025-11-19T08:08:01.0682482Z + validate_services http://192.168.122.213:10507/v1/docsum text docsum-vllm docsum-vllm '{"messages":"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.", "max_tokens":32, "language":"en", "stream":false}'
2025-11-19T08:08:01.0688600Z + local URL=http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:01.0689874Z + local EXPECTED_RESULT=text
2025-11-19T08:08:01.0690307Z + local SERVICE_NAME=docsum-vllm
2025-11-19T08:08:01.0690603Z + local DOCKER_NAME=docsum-vllm
2025-11-19T08:08:01.0691788Z + local 'INPUT_DATA={"messages":"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.", "max_tokens":32, "language":"en", "stream":false}'
2025-11-19T08:08:01.0694225Z ++ curl -s -o /dev/null -w '%{http_code}' -X POST -d '{"messages":"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.", "max_tokens":32, "language":"en", "stream":false}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:10.4631747Z + local HTTP_STATUS=200
2025-11-19T08:08:10.4632631Z + echo ===========================================
2025-11-19T08:08:10.4633476Z + '[' 200 -eq 200 ']'
2025-11-19T08:08:10.4634567Z + echo '[ docsum-vllm ] HTTP status is 200. Checking content...'
2025-11-19T08:08:10.4635475Z ===========================================
2025-11-19T08:08:10.4636326Z [ docsum-vllm ] HTTP status is 200. Checking content...
2025-11-19T08:08:10.4645311Z ++ curl -s -X POST -d '{"messages":"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.", "max_tokens":32, "language":"en", "stream":false}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:10.4647536Z ++ tee /home/sdp/opea-actions-runner/_work/GenAIComps/GenAIComps/tests/docsum-vllm.log
2025-11-19T08:08:20.0167266Z + local 'CONTENT={"id":"38068711ae9f07d41bfe363dac05dd98","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:20.0176933Z + echo '{"id":"38068711ae9f07d41bfe363dac05dd98","text":"' '\nTEI' is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, 'Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:20.0183014Z + echo '{"id":"38068711ae9f07d41bfe363dac05dd98","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}'
2025-11-19T08:08:20.0186349Z + grep -q text
2025-11-19T08:08:20.0188191Z {"id":"38068711ae9f07d41bfe363dac05dd98","text":" \nTEI is a toolkit for deploying and serving text embeddings and sequence classification models, allowing for high-performance extraction of popular models such as FlagEmbedding, Ember","prompt":"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.","output_guardrail_params":null}
2025-11-19T08:08:20.0189999Z + echo '[ docsum-vllm ] Content is as expected.'
2025-11-19T08:08:20.0190363Z + sleep 1s
2025-11-19T08:08:20.0190612Z [ docsum-vllm ] Content is as expected.
2025-11-19T08:08:21.0211988Z + echo 'Validate Chinese mode...'
2025-11-19T08:08:21.0213364Z Validate Chinese mode...
2025-11-19T08:08:21.0231080Z + validate_services http://192.168.122.213:10507/v1/docsum text docsum-vllm docsum-vllm '{"messages":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。", "max_tokens":32, "language":"zh", "stream":false}'
2025-11-19T08:08:21.0232637Z + local URL=http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:21.0233005Z + local EXPECTED_RESULT=text
2025-11-19T08:08:21.0233389Z + local SERVICE_NAME=docsum-vllm
2025-11-19T08:08:21.0233705Z + local DOCKER_NAME=docsum-vllm
2025-11-19T08:08:21.0234865Z + local 'INPUT_DATA={"messages":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。", "max_tokens":32, "language":"zh", "stream":false}'
2025-11-19T08:08:21.0236331Z ++ curl -s -o /dev/null -w '%{http_code}' -X POST -d '{"messages":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。", "max_tokens":32, "language":"zh", "stream":false}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:30.4520895Z + local HTTP_STATUS=200
2025-11-19T08:08:30.4521458Z + echo ===========================================
2025-11-19T08:08:30.4521832Z + '[' 200 -eq 200 ']'
2025-11-19T08:08:30.4522204Z + echo '[ docsum-vllm ] HTTP status is 200. Checking content...'
2025-11-19T08:08:30.4522634Z ===========================================
2025-11-19T08:08:30.4523084Z [ docsum-vllm ] HTTP status is 200. Checking content...
2025-11-19T08:08:30.4533203Z ++ curl -s -X POST -d '{"messages":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。", "max_tokens":32, "language":"zh", "stream":false}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:30.4535016Z ++ tee /home/sdp/opea-actions-runner/_work/GenAIComps/GenAIComps/tests/docsum-vllm.log
2025-11-19T08:08:39.2028562Z + local 'CONTENT={"id":"344c2ad5ad61a0dcc1e89e5b067e709b","text":"英特尔发布了新的处理器英特尔®至强®6性能核处理器（代号Granite Rapids），旨在为AI","prompt":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。","output_guardrail_params":null}'
2025-11-19T08:08:39.2029984Z + echo '{"id":"344c2ad5ad61a0dcc1e89e5b067e709b","text":"英特尔发布了新的处理器英特尔®至强®6性能核处理器（代号Granite' 'Rapids），旨在为AI","prompt":"2024年9月26日，北京——今日，英特尔正式发布英特尔®' 至强® 6性能核处理器（代号Granite 'Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。","output_guardrail_params":null}'
2025-11-19T08:08:39.2031649Z {"id":"344c2ad5ad61a0dcc1e89e5b067e709b","text":"英特尔发布了新的处理器英特尔®至强®6性能核处理器（代号Granite Rapids），旨在为AI","prompt":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。","output_guardrail_params":null}
2025-11-19T08:08:39.2035186Z + echo '{"id":"344c2ad5ad61a0dcc1e89e5b067e709b","text":"英特尔发布了新的处理器英特尔®至强®6性能核处理器（代号Granite Rapids），旨在为AI","prompt":"2024年9月26日，北京——今日，英特尔正式发布英特尔® 至强® 6性能核处理器（代号Granite Rapids），为AI、数据分析、科学计算等计算密集型业务提供卓越性能。","output_guardrail_params":null}'
2025-11-19T08:08:39.2041545Z + grep -q text
2025-11-19T08:08:39.2062063Z [ docsum-vllm ] Content is as expected.
2025-11-19T08:08:39.2063124Z + echo '[ docsum-vllm ] Content is as expected.'
2025-11-19T08:08:39.2063403Z + sleep 1s
2025-11-19T08:08:40.2092978Z + echo 'Validate truncate mode...'
2025-11-19T08:08:40.2094466Z Validate truncate mode...
2025-11-19T08:08:40.2100252Z + validate_services http://192.168.122.213:10507/v1/docsum text docsum-vllm docsum-vllm '{"messages":"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.", "max_tokens":32, "language":"en", "summary_type": "truncate", "chunk_size": 2000}'
2025-11-19T08:08:40.2106348Z + local URL=http://192.168.122.213:10507/v1/docsum
2025-11-19T08:08:40.2107459Z + local EXPECTED_RESULT=text
2025-11-19T08:08:40.2108654Z + local SERVICE_NAME=docsum-vllm
2025-11-19T08:08:40.2109602Z + local DOCKER_NAME=docsum-vllm
2025-11-19T08:08:40.2113589Z + local 'INPUT_DATA={"messages":"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.", "max_tokens":32, "language":"en", "summary_type": "truncate", "chunk_size": 2000}'
2025-11-19T08:08:40.2116224Z ++ curl -s -o /dev/null -w '%{http_code}' -X POST -d '{"messages":"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.", "max_tokens":32, "language":"en", "summary_type": "truncate", "chunk_size": 2000}' -H 'Content-Type: application/json' http://192.168.122.213:10507/v1/docsum
2025-11-19T08:58:53.6967062Z ++ stop_service
```

### Attachments

[logs_50254316498.zip](https://github.com/user-attachments/files/23650216/logs_50254316498.zip)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[Bug] `llm-docsum` test fail on vllm backend #1947

Priority

OS type

Hardware type

Installation method

Deploy method

Running nodes

What's the version?

Description

Reproduce steps

Raw log

Attachments

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[Bug] llm-docsum test fail on vllm backend #1947

Description

Priority

OS type

Hardware type

Installation method

Deploy method

Running nodes

What's the version?

Description

Reproduce steps

Raw log

Attachments

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

[Bug] `llm-docsum` test fail on vllm backend #1947