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, 07:00 UTC
main
[MRV2] Avoid recompilation of _gather_block_tables_kernel (#37645)
Woosuk Kwon committed 8 months ago
[Perf] Eliminate redundant SparseMatrix creation in gpt_oss_triton_kernels (#37683)
Xin Yang committed 8 months ago
[ROCm][CI] Fix accuracy for llama-nemotron-vl pooling tests (#37613)
Andreas Karatzas committed 8 months ago
Fix various config related issues for Transformers v5 (#37681)
Harry Mellor committed 8 months ago
[compile] Add compiled artifact counter for VLLM_USE_MEGA_AOT_ARTIFACT=1. (#37589)
Zhengxu Chen committed 8 months ago
[Metrics] Some small refactoring for better maintainability (#33898)
Martin Hickey committed 8 months ago
[compile] Fix aot test failures with torch 2.12. (#37604)
Zhengxu Chen committed 8 months ago
[Pixtral] Enable Pixtral language model support Eagle3 (#37182)
Rémi Delacourt committed 8 months ago
[Bugfix] Fix ConchLinearKernel channelwise quantization (group_size=-1) (#37329)
Matthias Gehre committed 8 months ago
[Bugfix] Reject channelwise quantization (group_size <= 0) in ExllamaLinearKernel (#37331)
Matthias Gehre committed 8 months ago
[ROCm] Enable wvSplitK skinny GEMM kernel for RDNA4/gfx1x decode (#34709)
L.B.R. committed 8 months ago
[ROCm][Quantization] make quark ocp mx dtype parser robust for weight-only quantization (#36232)
xuebwang-amd committed 8 months ago
[Model] Add LFM2-ColBERT-350M support (#37528)
Ilya Boytsov committed 8 months ago
Fix attribute error in `isaac_patch_hf_runner` (#37685)
Harry Mellor committed 8 months ago
[UX] Enable torch_profiler_with_stack (#37571)
Jee Jee Li committed 8 months ago
[Misc] Use logger.info_once for auto tool choice log message (#37661)
Chauncey committed 8 months ago
[Refactor] Relocate tests from tests/v1/entrypoints/ to tests/entrypoints/ (#37500)
Flora Feng committed 8 months ago
[Refactor] Move serve entrypoint tests under tests/entrypoints/serve/ (#37595)
Flora Feng committed 8 months ago
[ROCm][CI] Fix granite_speech test for gfx90a by selecting compatible attention backend (#37611)
Andreas Karatzas committed 8 months ago
[ROCm][CI] Remove deepep DBO tests on gfx90a (#37614)
Andreas Karatzas committed 8 months ago