[MoE Refactor][5/N] Isolate zero expert to LongCatFlash (#28891)

baonudesifeizhai committed 11 months ago

[XPU] enable fp8 online streaming quantization (#30944)

Yan Ma committed 11 months ago

[Bugfix] Read truncate_prompt_tokens from pooling_params in AsyncLLM.encode() (#31013)

Jeffrey Wang committed 11 months ago

[CI] FIx `fixture 'siglip_attention_config' not found` (#31053)

Lucas Wilkinson committed 11 months ago

[ROCm][CI/Build] Update ROCm dockerfiles (#30991)

Gregory Shtrasberg committed 11 months ago

[Bugfix] fix the alias bug of AttentionBackendEnum when register CUSTOM attention backend to vllm (#30869)

zejunchen-zejun committed 11 months ago

GLM-4.7 Tool Parser and Doc Update (#30876)

Yuxuan Zhang committed 11 months ago

[MoE Refactor][2/N] Use Modular Kernels for Fp8 (#30825)

Robert Shaw committed 11 months ago

[Bug] Fix `error 'Dynamo failed to run FX node with fake tensors` for Deepseek V3.2 (#31046)

Wentao Ye committed 11 months ago

[MoE Refactor][3/N] Deprecate cutlass block quant fp8 (b200) (#30990)

Robert Shaw committed 11 months ago

[BugFix] Fix TypeError: unhashable type: 'dict' when serving deepseek32 (#30924)

Lucas Wilkinson committed 11 months ago

[Refactor] Refactor for `DeepGemmQuantScaleFMT` using cache (#30898)

Wentao Ye committed 11 months ago

Make engine core client handshake timeout configurable (#27444)

Seiji Eicher committed 11 months ago

[Model] Add MiMo-V2-Flash support (#30836)

Zhonghua Deng committed 11 months ago

Update Pytorch version update docs (#30982)

Andrey Talman committed 11 months ago

[Quantization] fix marlin w8a8 check (#30961)

Jinzhen Lin committed 11 months ago

[CustomOp][Refactor] Extract common methods for ApplyRotaryEmb CustomOp (#31021)

Shanshan Shen committed 11 months ago

[Bugfix] [Kernel] Triton attention kernels: mask out V blocks that fall outside sliding window (#30887)

Thomas Parnell committed 11 months ago

[CPU][Bugfix] Fix ppc64le CPU build (#30871)

Nishidha Panpaliya committed 11 months ago

Enable aarch64 CPU performance benchmarks (#26494)

Li, Jiang committed 11 months ago