[Bugfix][Mooncake] Report failed remote KV loads to the scheduler (#50984)

BruceLoveDecimal committed about 1 month ago

[CI] Keep OTel bytecode out of mounted checkouts (#56766)

Kevin H. Luu committed about 1 month ago

[ROCm][Perf] W4A16: keep skinny GEMM zero-points packed 4-bit (#54965)

Matthias Gehre committed about 1 month ago

[Perf][GLM-5.3-Flash] Dense/masked-MHA sparse prefill for the NoPE (256, 0, 256) layout + skip the NoPE K concat (#55738)

Jared Wen committed about 1 month ago

[Bugfix][CPU] Fix DeepSeek-R1 (FP8 MLA + MoE) correctness on CPU backend (#56773)

Li, Jiang committed about 1 month ago

[CPU][Profiler] Group torch profiler tables by input shape when record_shapes is on (#56016)

Chinmay-Kulkarni-AMD committed about 1 month ago

[Perf][GLM-5.3-Flash] Use FlashKDA for KDA chunked prefill (1.7-3.8x faster than the Triton chunk path) (#55737)

Jared Wen committed about 1 month ago

[CI] Update entrypoints CI (#56763)

wang.yuqi committed about 1 month ago

[Performance] Add Triton kernel for Gemma3n sparse GELU (#48498)

JooHo Lee committed about 1 month ago

[Bugfix][Rust Frontend] Preserve selected-token logprob mode (#56406)

Alec committed about 1 month ago

[Bugfix] Fix --lora-modules name=path parsing when path contains '=' (#53353)

Hung Hoang (黄兴） committed about 1 month ago

[Fast Start] Use GPU uuid as socket folder identifier (#56669)

Isotr0py committed about 1 month ago

[Refactor] Normalize DeepSeek V4.1 model package naming (#56741)

Bugen Zhao committed about 1 month ago

[Bugfix][Beam Search] Respect skip_special_tokens during decoding (#56211)

Alexander Lee committed about 1 month ago

[2/2][Model Runner V2] FULL CUDA graph capture for microbatched steps (DBO) (#51700)

Anjie Hou committed about 1 month ago

[Frontend] Move grpc_server to launchers. (#56746)

wang.yuqi committed about 1 month ago

[CI] Collect GPU memory telemetry for MIG slices (#56747)

Kevin H. Luu committed about 1 month ago

[CI] Move smaller H200 workloads to 18GB MIG slices (#56695)

Kevin H. Luu committed about 1 month ago

[Bugfix][Multimodal] Validate base64 video payloads, matching image and audio (#54323)

Hotragn Pettugani committed about 1 month ago

Revert "[CI][XPU] Disable model runner V2 for XPU quantization test for some partially pre-quantized models" (#56394)

Chaojun Zhang committed about 1 month ago