LM/vLLM - Revision history

Tacoball: /* compose */

2026-08-31T10:51:37Z

compose

2026-08-31T10:50:34Z

compose

2026-08-31T10:48:27Z

← Older revision		Revision as of 10:48, 31 August 2026
Line 1:		Line 1:
			== compose ==

			<syntaxhighlight lang="yaml">
			services:
			vllm-xpu:
			image: vllm/vllm-openai-xpu:latest
			container_name: vllm-xpu-qwen38
			privileged: true
			network_mode: host
			shm_size: "32g"

			devices:
			- /dev/dri:/dev/dri

			volumes:
			- ${HOME}/.cache/huggingface:/root/.cache/huggingface

			environment:
			- HF_HOME=/root/.cache/huggingface
			- HF_TOKEN=${HF_TOKEN:-}
			- no_proxy=localhost,127.0.0.1
			- ZE_FLAT_DEVICE_HIERARCHY=COMPOSITE
			- ZE_AFFINITY_MASK=0
			- VLLM_WORKER_MULTIPROC_METHOD=spawn
			- PYTORCH_ALLOC_CONF=expandable_segments:True
			- VLLM_XPU_ENABLE_XPU_GRAPH=1

			ports:
			- "9000:9000"

			entrypoint: ["vllm", "serve"]
			command:
			- "SergiioB/Qwen3.8-27B-GPTQ-Int4-sym-G128-MTP-BF16"
			- "--served-model-name=qwen38"
			- "--port=9000"
			- "--host=0.0.0.0"
			- "--quantization=gptq"
			- "--dtype=float16"
			- "--max-model-len=131072"
			- "--gpu-memory-utilization=0.88"
			- "--kv-cache-dtype=fp8"
			- "--max-num-seqs=1"
			- "--max-num-batched-tokens=8192"
			- "--trust-remote-code"
			- "--enable-auto-tool-choice"
			- "--tool-call-parser=qwen3_xml"
			- "--reasoning-parser=qwen3"
			- "--limit-mm-per-prompt={\"image\":10}"
			- "--speculative-config={\"method\":\"mtp\",\"num_speculative_tokens\":4}"

			restart: no
			</syntaxhighlight>

	== docker.io ==		== docker.io ==

2026-08-16T17:35:17Z

docker.io

← Older revision		Revision as of 17:35, 16 August 2026
Line 4:		Line 4:
	sudo apt install docker.io		sudo apt install docker.io
	sudo docker pull intel/vllm		sudo docker pull intel/vllm

			docker run -t -d --shm-size 10g --net=host --ipc=host --privileged \
			-v /dev/dri/by-path:/dev/dri/by-path --name=vllm-test \
			--device /dev/dri:/dev/dri --entrypoint=/bin/bash intel/vllm:latest
	</syntaxhighlight>		</syntaxhighlight>

2026-08-14T16:05:17Z

2026-08-14T03:05:53Z

TODO

← Older revision		Revision as of 03:05, 14 August 2026
Line 39:		Line 39:
	== TODO ==		== TODO ==

			* https://hub.docker.com/r/intel/vllm
	* https://hub.docker.com/u/intel/?page=1&search=vllm		* https://hub.docker.com/u/intel/?page=1&search=vllm
	* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements		* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements

2026-08-14T03:03:14Z

TODO

← Older revision		Revision as of 03:03, 14 August 2026
Line 39:		Line 39:
	== TODO ==		== TODO ==

			* https://hub.docker.com/u/intel/?page=1&search=vllm
	* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements		* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements
	* https://www.pugetsystems.com/labs/articles/intel-arc-pro-b70-multi-gpu-ai-inference-performance/		* https://www.pugetsystems.com/labs/articles/intel-arc-pro-b70-multi-gpu-ai-inference-performance/

2026-08-14T02:56:08Z

TODO

← Older revision		Revision as of 02:56, 14 August 2026
Line 44:		Line 44:
	* https://sergiiob.dev/posts/intel-arc-b70-vllm-initial-mxfp4-test/		* https://sergiiob.dev/posts/intel-arc-b70-vllm-initial-mxfp4-test/
	* https://forum.level1techs.com/t/intel-b70-launch-unboxed-and-tested/247873		* https://forum.level1techs.com/t/intel-b70-launch-unboxed-and-tested/247873
			* https://huggingface.co/blog/MatrixYao/intel-gpu

2026-08-14T02:55:17Z

TODO

← Older revision		Revision as of 02:55, 14 August 2026
Line 43:		Line 43:
	* https://github.com/Hal9000AIML/arc-pro-b70-inference-setup-ubuntu-server		* https://github.com/Hal9000AIML/arc-pro-b70-inference-setup-ubuntu-server
	* https://sergiiob.dev/posts/intel-arc-b70-vllm-initial-mxfp4-test/		* https://sergiiob.dev/posts/intel-arc-b70-vllm-initial-mxfp4-test/
			* https://forum.level1techs.com/t/intel-b70-launch-unboxed-and-tested/247873

2026-08-14T02:50:39Z

TODO

← Older revision		Revision as of 02:50, 14 August 2026
Line 40:		Line 40:

	* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements		* https://docs.vllm.ai/en/stable/getting_started/installation/gpu/#requirements
			* https://www.pugetsystems.com/labs/articles/intel-arc-pro-b70-multi-gpu-ai-inference-performance/
			* https://github.com/Hal9000AIML/arc-pro-b70-inference-setup-ubuntu-server
			* https://sergiiob.dev/posts/intel-arc-b70-vllm-initial-mxfp4-test/