A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 6, 12:00 UTC
main
[Bugfix] Fix base64 JPEG video frames returning empty metadata (#37301)
Yufeng He committed 8 months ago
[NIXL][Bugfix] metrics & testing minor bug (#36051)
Andy Lo committed 8 months ago
[Model] Enable LoRA support for tower and connector in H2OVL (#31696)
Shwetha Poojary committed 8 months ago
[Bugfix] Fix EP weight filter breaking EPLB and NVFP4 accuracy (#37322)
Elvir Crnčević committed 8 months ago
[Build] Bump python openai version (#32316)
Chauncey committed 8 months ago
[CI] Fix PaddleOCR-VL HF test failure due to create_causal_mask API rename (#37328)
Andreas Karatzas committed 8 months ago
[docs] Add docs for new RL flows (#36188)
Aaron Hao committed 8 months ago
fix(glm47): improve tool call parsing and content normalization (#37386)
Karan Bansal committed 8 months ago
[LoRA] Make LoRA respect `language_model_only` (#37375)
Jee Jee Li committed 8 months ago
[Bugfix] Avoid OpenMP thread reallocation in CPU torch compile (#37391)
Li, Jiang committed 8 months ago
[kv_offload+HMA][0/N]: Support block-level preemption handling (#34805)
Or Ozeri committed 8 months ago
[XPU] skip unsupported ut and update test_nixl_connector (#37179)
liuzhenwei committed 8 months ago
[responsesAPI] parser.extract_response_outputs can take in token IDs (#37130)
Andrew Xia committed 8 months ago
[CI] Stabilize test_cpu_offloading by waiting for async offload before cache reset (#37335)
Andreas Karatzas committed 8 months ago
[ROCm][CI] Add ROCM_EXTRA_ARGS to audio_in_video test server fixture (#37349)
Andreas Karatzas committed 8 months ago
gxd3 committed 8 months ago
[Perf] Enable dual stream execution of input projection for Qwen3 (#36795)
Xin Yang committed 8 months ago
[ROCm][CI] Skip trtllm kvfp8 dequant tests on ROCm (#37330)
Andreas Karatzas committed 8 months ago
[Performance] Add --enable-ep-weight-filter CLI option (#37351)
Roy Wang committed 8 months ago
[Kernel][Helion] [16/N] Refactor register_kernel API to be more Dynamo-friendly (#36705)
Yanan Cao committed 8 months ago