[ROCm][MoE] Pad the AITER MoE intermediate size at allocation time, and round the expert-group count to a kernel that exists (#55368)

Sergey Shlyapnikov committed 21 days ago

[ROCm][Perf][GLM-5.3-Flash] "Fit kpool top-k indices to AITER" with a single Triton kernel (#58008)

Simon Danielsson committed 21 days ago

[Refactor] Share workspace and model runner init between GPU and XPU workers (#59200)

aoshen02 committed 21 days ago

[Security] Bump pyjwt and rand for remaining GHSAs (#59427)

Juan Pérez de Algaba committed 21 days ago

[Feat] Support PP with PCP in GPU Model Runner V2 (#59139)

Qiu Chunshuo committed 21 days ago

[Bugfix][Core] Schedule encoder-only prompts larger than one step (#59029)

Giulio De Pasquale committed 21 days ago

[Test] Skip IPC weight-transfer test on XPU platforms (#59379)

Chaojun Zhang committed 21 days ago

[CI/Build] Fix pre-commit (#59426)

Cyrus Leung committed 21 days ago

[KV-Offloading][TP] : Expand replicated_layout detection to multi-group MLA (#57652)

Varun Sundar Rabindranath committed 21 days ago

[Feature][Rust Frontend] Add Shutdown control RPC (#59316)

JulienDarve committed 21 days ago

[Docs] Reinstate docs build gate for PRs (#59411)

Harry Mellor committed 21 days ago

[MM][Mistral] Add compile support for Pixtral vision encoders (#57168)

Jakub Zakrzewski committed 21 days ago

[CPU] Conv1d optimised kernel for aarch64 (#54093)

almayne committed 21 days ago

[XPU] Dispatch nn.LayerNorm to fused SYCL kernel via CustomOp (#57172)

Michał Ganczarenko committed 21 days ago

[Security] Bump remaining Dependabot packages (excl. ignored) (#59315)

Juan Pérez de Algaba committed 21 days ago

[Security] Authenticate shared-memory multimodal cache handles (#59357)

Clinton Thomas committed 21 days ago

[Hardware][Power] Enable W4A16 (AWQ & GPTQ) quantization on POWER10 using VSX (#59149)

Rukhaiya2004 committed 21 days ago

docs: add gemma-2, SmolLM2, Qwen2.5-Coder, Llama-3.2-1B to batch invariance tested models (#54441)

yuvalluria committed 21 days ago

[MM] Keep device input normalization fused when encoder compilation is enabled (#59195)

Jakub Zakrzewski committed 21 days ago

[MyPy][3/N] Fix MyPy errors in test groups (part 3) (#55939)

Taimys committed 21 days ago