[Pooling] Preserve BERT-family heads for raw logits (#57664)

Taneem Ibrahim committed 26 days ago

[AuxOutput] Only require Model Runner V2 on GPU platform (#58205)

Linkun committed 26 days ago

[Bugfix] Don't drop the rest of the allocator config when toggling expandable segments (#57982)

Oxana Korzh committed 26 days ago

[watermarking] golden tests for backwards compatibility (#56809)

Raphaël Rialland committed 26 days ago

[Bugfix] Default missing detail for Responses API input images (#57241)

yzong-rh committed 26 days ago

[Perf][Frontend] Defer reasoning usage recounts for non-continuous chat streams (#56067)

Cheng Rui committed 26 days ago

[Bugfix][Frontend] Document 404 response for `/generative_scoring` (#58788)

Nick Hill committed 26 days ago

[Bugfix][Frontend] Respect max_output_tokens in the Harmony tool-call loop (#58551)

Denis Ermakov committed 26 days ago

[Feature][Frontend] Add granite_thinking_parser reasoning parser for Granite 4.2 (#55957)

Yousaf committed 26 days ago

[ROCm] Bump torch 2.13, triton 3.8, torchaudio, torchvision (#50605)

Rohan Potdar committed 26 days ago

[ROCm][CI] Run the MLA attention+quant fusion test on ROCm (#58717)

stefankoncarevic committed 26 days ago

[Bugfix][CI] Report subprocess test skips as skips, not passes (#58701)

stefankoncarevic committed 26 days ago

[CI] [MRV2] Restore MRV2 pp dp coverage (#57735)

Taneem Ibrahim committed 26 days ago

[ROCm][CI] Expand single-GPU coverage on MI355 DPX (#57599)

Sheral Kumar committed 26 days ago

[Perf] DiffusionGemma: one-pass sampler statistics kernel (#58226)

Matt Mastracci committed 26 days ago

[ROCm][CI] Test AMD DeepSeek V4 MoE routing against a PyTorch reference (#58740)

stefankoncarevic committed 26 days ago

[ROCm][CI] Add quantized MoE serving test for gfx950 (#58748)

stefankoncarevic committed 26 days ago

[ROCm][CI] Cover the AITER MQA logits dispatch on gfx950 (#58724)

stefankoncarevic committed 26 days ago

[ROCm][Kimi-K3] Optimize low-concurrency speculative KDA (#58045)

jiacao-amd committed 26 days ago

[Bugfix][KV Connector] Reap expired NIXL leases behind a heartbeated head (#58292)

GokayAI committed 26 days ago