[Bugfix][Multimodal] Encode text in mixed CLIP/SigLIP pooling batches (#53165)

Prudhvi Vuda committed 2 months ago

[MISC] Cleanup deprecated parameters (#53559)

Jee Jee Li committed 2 months ago

[Core] drop duplicate VLLM_USE_DEEP_GEMM check (#48687)

Ning Xie committed 2 months ago

[Kernel][Perf] Per-architecture tuned configs for batch-invariant persistent matmul (~3x decode kernels on RTX 4090D/H20) (#53247)

Lio Einaudi committed 2 months ago

[Kimi K3 Refactor] Add `UnfinalizedMoEOutput` proto following up for #53152 (#53310)

Wentao Ye committed 2 months ago

[Bugfix][Kimi K3] Skip absent metadata during CUDA graph profiling (#53581)

Kevin H. Luu committed 2 months ago

[Bugfix][Spec Decode] Reapply group geometry for FlashAttention metadata (#53336)

Misha Goin committed 2 months ago

[Frontend] Move run_batch.py out openai folder (#53500)

wang.yuqi committed 2 months ago

[Bugfix][DCP] Handle sparse MLA metadata after DCP Manager refactor (#52377)

cjackal committed 2 months ago

[MM] Cache common token sequences (#53560)

Cyrus Leung committed 2 months ago

[Bugfix] Release worker RPC payload before next dequeue (#51979)

石皮幼鸟 committed 2 months ago

[Docs][Security] Document multimodal media UUID security implications (#53582)

Juan Pérez de Algaba committed 2 months ago

[Pooling] Improve BGE-M3 sync pooling throughput by up to 3.13% (#53464)

Taneem Ibrahim committed 2 months ago

Add Cohere ChatV2 render endpoint (#53219)

andrewbcohere committed 2 months ago

[Bugfix][Frontend] Fix run_batch upload retrying on success and unawaited error body (#50588)

Tan Chao committed 2 months ago

[Bugfix][Kernel] Handle kernel block sizes in V2 DCP slot mapping (#51031)

Nicolò Lucchesi committed 2 months ago

Fix MNNVL Lamport mailbox publication and cleanup (#53000)

Ace Eldeib committed 2 months ago

[Bugfix] Reuse CUDA streams in packed weight transfer to cap reserved-memory waste (#52951)

Quentin Gallouédec committed 2 months ago

[Xeon][doc]add Xeon recipes into table (#53226)

Louie Tsai committed 2 months ago

[Model Runner V2] Reserve CUDA graph memory (#53306)

Nick Hill committed 2 months ago