[Bugfix][Pooling] Handle multimodal cache misses without crashing (#58975)

Chaoying committed 19 days ago

[Bugfix][Multimodal] Fix reference counting of the SHM processor cache (#59695)

tenotek committed 19 days ago

[Bugfix][Quark] Pass grouped-routing arguments to OCP MX monolithic kernels (#59752)

aoshen02 committed 19 days ago

[Perf] Keep non-speculative GDN decode on the standard path (#59735)

DimensionSTP committed 19 days ago

[Model] Add Qwen4Exp to the Qwen GDN Triton warmup (#56742)

SeongJun Lee committed 19 days ago

[Docs] Fix stale W4A16 NVFP4 default kernel in ModelOpt docs (#59135)

Rajeev Warrier committed 19 days ago

[CI][Kimi-K3] Test prefix cache reuse with KV offload, P/D and DCP (#59229)

Jiangyun Zhu committed 19 days ago

[Bugfix][Quantization] Detect NVFP4 in ModelOpt mixed-precision checkpoints (#56050)

Stefano Castagnetta committed 19 days ago

[Test] Re-enable Ovis2.5 and Ovis2.6-MoE vLLM tests on Transformers v5 (#59763)

Abhay Shukla committed 19 days ago

[XPU][Kernel] Tune Triton W8A8 block-FP8 GEMM for Intel B70 (#56063)

pmanczak committed 19 days ago

[CI] Replace shellcheck-suppressed patterns flagged in #52572 with clean equivalents (#53341)

Turner Jabbour committed 19 days ago

[Model] Migrate GPT-NeoX, Phi, Seed-OSS and Jais2 to the Transformers modeling backend (#59701)

Harry Mellor committed 19 days ago

[Misc] Remove code paths for Transformers < 5.16.1 (#59762)

Harry Mellor committed 19 days ago

[Bugfix][LoRA] Fall back for high-rank MoE LoRA (#55161)

xiaoyu-xyz committed 19 days ago

[ROCm][Perf][GLM-5.3-Flash] Add AITER topk backend for decodes (#58167)

Simon Danielsson committed 19 days ago

[ROCm][Perf][GLM-5.3-Flash] Remove redundant copy after ragged sparse MLA (#58569)

Simon Danielsson committed 19 days ago

[Bugfix][GLM-5.3] Size the image encoder cache from the exact token ceiling (#59565)

Yufeng He committed 19 days ago

[Bugfix][Spec Decode] Qualify Transformers backend attention layer names with the model prefix (#59293)

Farhad Al-Amin committed 19 days ago

[Feature] Offload KV-init runtime state on sleep (#59158)

aoshen02 committed 19 days ago

[Security] Accept zero-sum DeepSeek-OCR pixel tensors (#59417)

Juan Pérez de Algaba committed 19 days ago