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 13, 15:45 UTC
main
[CI/Build] Loosen STT LoRA Translate Check (Flaky Test) (#28247)
Alex Brooks committed about 1 year ago
[Multimodal][torch.compile] Add compilation config field for turning off ViT/MM compile (#28242)
Lucas Kabela committed about 1 year ago
Speed up mm processor kwargs per request by spliting dynamic and static kwargs (#26483)
Junhong Liu committed about 1 year ago
Fix failing test for CRadio (#27738)
Eugene Khvedchenya committed about 1 year ago
[BugFix] Fix FusedMoELoRA + ModularKernel Integration (#28237)
Varun Sundar Rabindranath committed about 1 year ago
[Test] Add non-MoE DP test coverage (#28235)
Matthew Bonanni committed about 1 year ago
[Doc]: Make extraInit containers fully configurable in helm chart (#27497)
Fang Han committed about 1 year ago
[ROCm] triton fp8 kernel (#27058)
Aleksandr Malyshev committed about 1 year ago
Vico Chu committed about 1 year ago
Add llama 4 scaling support (#28145)
Julien Denize committed about 1 year ago
[Structured outputs] Upgrade llguidance to 1.3.0 (#28039)
Andy Lo committed about 1 year ago
CODEOWNERS: Add myself as reviewer on security docs (#28216)
Russell Bryant committed about 1 year ago
[Frontend] Fix logging format when enable response logging (#28049)
Roy Wang committed about 1 year ago
[HARDWARE][CPU] Add Option for Disabling Binding to Specific CPU Cores (#27953)
StanHatko committed about 1 year ago
[Doc] Add Arm CPUs are on the list of supported targets in vLLM (#26018)
Milos Puzovic committed about 1 year ago
[Kernel][Model] Tune fused_moe Triton configs for MiniMax-M2 on H100 (#28200)
Eric Yue committed about 1 year ago
[Docs] Switch to directory style URLs (#28058)
Harry Mellor committed about 1 year ago
Disable nm-testing models with issues in CI (#28206)
Michael Goin committed about 1 year ago
[CPU]Improve cpu fused moe perf (#27244)
xiangze-arm committed about 1 year ago
[Frontend] OpenAI Responses API supports Tool/Function calling - non-harmony (#26874)
Chauncey committed about 1 year ago