[Attention][Cleanup] Remove tree attention (#42121)

Lucas Wilkinson committed 6 months ago

[XPU] remove q/k/v force contiguous for flash_attn (#40356)

Yan Ma committed 6 months ago

[feat] Add explicit `/start_weight_update` and `/finish_weight_update` APIs for weight transfer (#39212)

Sumanth R Hegde committed 6 months ago

Use `pre-commit / pre-run-check` to gate docs build too (#42053)

Harry Mellor committed 6 months ago

[Bugfix][KV Transfer] Reject NixlConnector + expandable_segments:True (#41237)

Roy Wang committed 6 months ago

[Bugfix] Fix FlashInfer CUTLASS MXFP4-MXFP8 MoE by restoring swizzled scale (#42089)

Michael Goin committed 6 months ago

[CUDA][CUTLASS] Enable cutlass scaled mm for non-compatible sizes (#41868)

John Calderon committed 6 months ago

use HIP_VERSION variables to guard against duplicate atomicAdd definitions (#41802)

pmaybank committed 6 months ago

[CI] Narrow misc.yaml source dependencies (#42059)

Kevin H. Luu committed 6 months ago

[CI] Narrow engine.yaml source dependencies (#42055)

Kevin H. Luu committed 6 months ago

[CI] Narrow Platform Tests (CUDA) source dependencies (#42054)

Kevin H. Luu committed 6 months ago

[CI] Narrow pytorch.yaml compile job source dependencies (#42057)

Kevin H. Luu committed 6 months ago

[Bugfix][Gemma4] Fix infinite loop and array boundary issues in tool parser (#41991)

David Oy committed 6 months ago

[ROCm][Perf] Add Fused Shared Expert (FSE) support for Qwen3-Next (#39280)

Nico Holmberg committed 6 months ago

[CI][Elastic EP] Fix Elastic EP Scaling Test Failure (#41792)

haosdent committed 6 months ago

[Models][Gemma3/Gemma4] Support hidden_act variants in gated MLP (#40588)

Hiroaki Mikami committed 6 months ago

[Docs] Fix OpenAI batch model argument examples (#42066)

Ethan Feng committed 6 months ago

[CI][XPU] Skip fork-dependent logits processor test (#42013)

liuzhenwei committed 6 months ago

[CI][Examples][RLHF] Disable async scheduling in rlhf_async_new_apis (#42042)

haosdent committed 6 months ago

Make docs environment deterministic (#41926)

Harry Mellor committed 6 months ago