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
[doc] backward compatibility for 0.6.4 (#11359)
youkaichao committed almost 2 years ago
[misc] add early error message for custom ops (#11355)
youkaichao committed almost 2 years ago
[Bugfix] Fix spec decoding when seed is none in a batch (#10863)
Wallas Henrique committed almost 2 years ago
[Misc] Add tqdm progress bar during graph capture (#11349)
Michael Goin committed almost 2 years ago
[ci][gh200] dockerfile clean up (#11351)
youkaichao committed almost 2 years ago
[Bugfix][Hardware][POWERPC] Fix auto dtype failure in case of POWER10 (#11331)
Akash kaothalkar committed almost 2 years ago
[CI] Adding CPU docker pipeline (#11261)
Yuan committed almost 2 years ago
[Misc] Remove unused vllm/block.py (#11336)
yangzhibin committed almost 2 years ago
[Bugfix] Fix broken CPU compressed-tensors test (#11338)
Isotr0py committed almost 2 years ago
[Misc] Clean up and consolidate LRUCache (#11339)
Cyrus Leung committed almost 2 years ago
[Model] Refactor Qwen2-VL to use merged multimodal processor (#11258)
Isotr0py committed almost 2 years ago
[V1] Fix multimodal profiling for `Molmo` (#11325)
Roger Wang committed almost 2 years ago
[Model] Add JambaForSequenceClassification model (#10860)
Yehoshua Cohen committed almost 2 years ago
[Bugfix] Cleanup Pixtral HF code (#11333)
Cyrus Leung committed almost 2 years ago
[Feature] Add load generation config from model (#11164)
Yanyi Liu committed almost 2 years ago
[misc] benchmark_throughput : Add LoRA (#11267)
Varun Sundar Rabindranath committed almost 2 years ago
[Misc] Optimize ray worker initialization time (#11275)
Rui Qiao committed almost 2 years ago
[Kernel] Refactor Cutlass c3x (#10049)
Varun Sundar Rabindranath committed almost 2 years ago
[VLM] Merged multimodal processor for Qwen2-Audio (#11303)
Cyrus Leung committed almost 2 years ago
[V1] Simplify prefix caching logic by removing `num_evictable_computed_blocks` (#11310)
Chen Zhang committed almost 2 years ago