[XPU][Tests] Enable test_per_token_group_quant_int8 on XPU (#55681)

pmanczak committed about 1 month ago

[Bugfix][Core] Retire Mamba states across null gaps (#55450)

lucamotz committed about 1 month ago

[ROCm][CI] Accept `base-v2-preview` images during content hash lookup for ROCm base images (#56356)

Micah Williamson committed about 1 month ago

[Docs] Add Nebius Serverless AI deployment guide (#56414)

Alexander committed about 1 month ago

[Attention] Extend XQA decode support on SM90 (#50439)

Andrii Skliar committed about 1 month ago

[Bugfix] Fix FlashInfer KV sharing with omitted K/V (#55864)

Zhicheng Liu committed about 1 month ago

[Frontend][last/N] Move all non-OpenAI content out of the OpenAI folder. (#56369)

wang.yuqi committed about 1 month ago

[Bugfix][MM] Fix swapped H/W in dummy video profiling inputs (#56385)

Hung Hoang (黄兴） committed about 1 month ago

[Perf][ROCm] Add AITER custom AG/RS (DP only) (#48247)

Simon Danielsson committed about 1 month ago

[PCP][Spec Decode] Adds PCP support for single-module MTP and replicated DSpark. (#56107)

Lucas Wilkinson committed about 1 month ago

[EC] Automatically enable embedding inputs on EC/KV consumers (#56379)

Tianyu Guo committed about 1 month ago

[Bugfix][LoRA] Use stored rsLoRA scaling factor in MoE expert packing (#55548)

Kushal committed about 1 month ago

[Bugfix][Mooncake] Fix heterogeneous PP transfer completion (#56033)

Yicong Wang committed about 1 month ago

[Feature][SimpleCPU] Load fine-grained hybrid prefix hits (#54736)

YukioZzz committed about 1 month ago

[ROCm][Bugfix] Route GLM-5.3-Flash MTP through ragged sparse MLA (#55239)

James E T Smith committed about 1 month ago

[Bugfix][Core] Stop zero-progress preemption cascades for deferred KV frees (#49675)

焦恺文 committed about 1 month ago

[XPU] Add forward_xpu to Mixer2RMSNormGated and FusedRMSNormGated (#54968)

Marceli Fylcek committed about 1 month ago

[CI/Build] Pin HyperCLOVAX V2 test model revision (#56335)

Tyler Michael Smith committed about 1 month ago

[Bugfix] Set stop_sequence explicitly in streaming message_delta event (#55325)

Colin McNamara committed about 1 month ago

[ROCm][Bugfix][Perf] Tune multi-stream shared experts use; wvSplitKrc fixes (#56098)

Matt committed about 1 month ago