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 4, 20:00 UTC
main
[MyPy] Enable mypy for `vllm/model_executor/layers/` (#40159)
Martin Hickey committed 7 months ago
[Multimodal] Add PyAV video backend for concurrent video decoding (#39986)
Jaseel Muhammad committed 7 months ago
rasmith committed 7 months ago
[Bugfix] Fix W4A8_FP8 MoE tp>1 correctness and view() TypeError (#40310)
EdalatiAli committed 7 months ago
[NIXL][XPU]Fix nixl import on XPU (#40430)
Soila Kavulya committed 7 months ago
rasmith committed 7 months ago
test: add nan/inf clamp regression test for fused_topk_bias (#40553)
Jhao-Ting Chen committed 7 months ago
[Bugfix] avoid warmup if text only expectation in multi_modal run (#40409)
Khushali Desai committed 7 months ago
TJian committed 7 months ago
[Startup][UX] Enable CUDAGraph memory profiling by default (#38284)
Matthew Bonanni committed 7 months ago
[MoE Refactor] Add more MoE layer tests (#39349)
bnellnm committed 7 months ago
[MoE Refactor] Remove SharedFusedMoE class (#35782)
bnellnm committed 7 months ago
[Refactor] Remove unused param (#39750)
Wentao Ye committed 7 months ago
[Perf] Optimize batch invariant with fused rms norm, 2.1% E2E latency improvement (#40413)
Wentao Ye committed 7 months ago
Jakub Zakrzewski committed 7 months ago
[Performance] Add is_reasoning_end_streaming() override to GptOssReasoningParser (#35745)
Fergus committed 7 months ago
[Bugfix] Fix spec decode test failures on Blackwell (SM100+) (#39546)
Rishi Puri committed 7 months ago
Zijing Liu committed 7 months ago
Revert "[Misc] Move `pyav` and `soundfile` to common requirements" (#40276)
Isotr0py committed 7 months ago
Add new tp plan styles to the Transformers modelling backend (#40467)
Harry Mellor committed 7 months ago