admin/vLLM_Head_Run

Go to file

JK Woo 5063aa98d8 Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

nccl-tests @ 036a7eda0d

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

.env

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

ANTIGRAVITY_RULES.md

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

api_key

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

create_gitea_repo.py

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

Dockerfile

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

kill-vllm.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

load-test-vllm_4.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

load-test-vllm_8.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

load-test-vllm_16.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

README.md

Initial commit

2026-06-26 22:43:32 +09:00

run-head.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

run-head.sh.bak

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-ray_nccl_log.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-ray.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-vllm_16k_seq8_stable.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-vllm-A.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-vllm.sh

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

start-vllm.sh.bak

Initial commit of vLLM Head Run project files

2026-06-26 14:05:01 +00:00

README.md

vLLM_Head_Run

"vLLM Head Run" 프로젝트는 torchrun 기반으로 통합된 vLLM OpenAI API 서버를 프로덕션 환경에서 효율적으로 배포하고 운영하기 위한 스크립트 및 관련 설정을 제공합니다. ConnectX-7(RoCE v2) 인프라 성능 최적화, 분산 학습 및 추론 성능 극대화, 그리고 ANTIGRAVITY_RULES.md에 명시된 개발 원칙 준수를 목표로 합니다.

Description

"vLLM Head Run" 프로젝트는 torchrun 기반으로 통합된 vLLM OpenAI API 서버를 프로덕션 환경에서 효율적으로 배포하고 운영하기 위한 스크립트 및 관련 설정을 제공합니다. ConnectX-7(RoCE v2) 인프라 성능 최적화, 분산 학습 및 추론 성능 극대화, 그리고 ANTIGRAVITY_RULES.md에 명시된 개발 원칙 준수를 목표로 합니다.

Readme 5.2 MiB