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 16, 04:30 UTC
main
[Misc] Add assertion and helpful message for marlin24 compressed models (#11388)
Dipika Sikka committed almost 2 years ago
[Bugfix] torch nightly version in ROCm installation guide (#11423)
Yuan Tang committed almost 2 years ago
[Bugfix] Fix CFGGuide and use outlines for grammars that can't convert to GBNF (#11389)
Michael Goin committed almost 2 years ago
mypy type checking for vllm/worker (#11418)
Lucas Tucker committed almost 2 years ago
[misc][perf] remove old code (#11425)
youkaichao committed almost 2 years ago
[CI] Unboock H100 Benchmark (#11419)
Simon Mo committed almost 2 years ago
[Bugfix] Fix fully sharded LoRAs with Mixtral (#11390)
Jason T. Greene committed almost 2 years ago
[cd][release] fix race conditions (#11407)
youkaichao committed almost 2 years ago
[cd][release] add pypi index for every commit and nightly build (#11404)
youkaichao committed almost 2 years ago
[CI] Fix flaky entrypoint tests (#11403)
Roger Wang committed almost 2 years ago
[Bugfix] Fix issues for `Pixtral-Large-Instruct-2411` (#11393)
Roger Wang committed almost 2 years ago
[V1] Make AsyncLLMEngine v1-v0 opaque (#11383)
Ricky Xu committed almost 2 years ago
[Bugfix] update should_ignore_layer (#11354)
George committed almost 2 years ago
[V1][Bugfix] Skip hashing empty or None mm_data (#11386)
Woosuk Kwon committed almost 2 years ago
Add ray[default] to wget to run distributed inference out of box (#11265)
Jiaxin Shan committed almost 2 years ago
[doc] explain nccl requirements for rlhf (#11381)
youkaichao committed almost 2 years ago
[Bugfix] Don't log OpenAI field aliases as ignored (#11378)
Michael Goin committed almost 2 years ago
[Core] Loading model from S3 using RunAI Model Streamer as optional loader (#10192)
omer-dayan committed almost 2 years ago
[CI/Build] fix pre-compiled wheel install for exact tag (#11373)
Daniele committed almost 2 years ago
[V1] Fix profiling for models with merged input processor (#11370)
Roger Wang committed almost 2 years ago