[Test][Qwen3-VL] Cover compiled DeepStack input contract (#53529)

maithilijoshi20 committed about 2 months ago

[KV Offload] Preserve KV event metadata until final residency removal (#52068)

Moein Khazraee committed about 2 months ago

[Frontend] Move engine/protocol.py out openai folder (#54492)

wang.yuqi committed about 2 months ago

[Bugfix][KVOffload] P2P tier declares REQUEST_LEVEL on the producer leg (#52912)

liranschour committed about 2 months ago

[Bugfix][SM120] DSv4: pass contiguous C128A decode topk indices on SM120 (#53574)

Gabriel Wu committed about 2 months ago

[Bugfix][Frontend] Truncate prompt_is_token_ids with the prompt (#54509)

Hotragn Pettugani committed about 2 months ago

[Bugfix][MM] Fix MiniCPM-o image processor reuse on Transformers v5 (#54501)

Andreas Karatzas committed about 2 months ago

[Renderer] Shutdown the renderer properly. (#52124)

wang.yuqi committed about 2 months ago

[Quantization][Autoround][XPU] Support AutoRound MXFP8 MoE models (#51248)

jl9876 committed about 2 months ago

[CPU] Support FP16/BF16 persisted GDN state on AMX (#52191)

Tianmu Li committed about 2 months ago

[XPU] Ensure unquantized linear weight is N-contiguous (#53536)

zofia committed about 2 months ago

[Bugfix][CPU] Fix several bugs (#54042)

Li, Jiang committed about 2 months ago

[Bugfix][DCP] Fix NVIDIA DeepSeek-V3.2 / GLM-5.2 fused attention (#50005)

Yunxiao Ning committed about 2 months ago

[KV Offload] Forward ownership in KV cache events (#52067)

Moein Khazraee committed about 2 months ago

[Kernel] Retire the DSv3 router GEMM CUDA kernel (#54040)

Jee Jee Li committed about 2 months ago

[Performance] Optimize Dots3 NOTE runtime (#53517)

范裕达 committed about 2 months ago

[Bugfix][Multimodal] Honor modality-scoped mm_processor_kwargs (#53808)

Prudhvi Vuda committed about 2 months ago

[Frontend] Forward cache salt for content parts (#54315)

eligotts committed about 2 months ago

[Bugfix][Frontend] Truncate prompt_token_offsets with the prompt (#54407)

Hotragn Pettugani committed about 2 months ago

[Model] Support Qwen3.8-Flash-Next (#53896)

Harry Huang committed about 2 months ago