[Bugfix] Fix race condition in KVBlockZeroer (#48085)

Benjamin Chislett committed 4 months ago

Add Intel XPU Docker release pipeline (#47880)

wenjun liu committed 4 months ago

Remove TeleChatForCausalLM (#47989)

Tiezhen WANG committed 4 months ago

[Bugfix] Fix Qwen3-ASR transcription streaming postprocessing (#42478)

JooHo Lee committed 4 months ago

[CPU] Fix Qwen-Next SSM type for AMX GDN (#48073)

Li, Jiang committed 4 months ago

[KV Offloading] Add free block iterator for CPU offload scheduling (#47849)

Chauncey committed 4 months ago

[XPU][LoRA] Fix torch.compile DEVICE_LOST by avoiding view-mutation in LoRA shrink (#47944)

Chaojun Zhang committed 4 months ago

[Bugfix][ROCm] Change AttentionCGSuppoort in TritonMLA to UNIFORM_SINGLE_TOKEN_DECODE (#47144)

music-dino committed 4 months ago

[ROCm][CI][MoE] Fix double-transpose of fused w3 expert weights (#47874)

stefankoncarevic committed 4 months ago

[KVConnector][NIXL] Support pipeline-parallel prefill in push mode (#45880)

qizixi committed 4 months ago

[Bugfix] Use int8 workspace for FlashInfer MLA decode (#48046)

Nick Hill committed 4 months ago

Fix embed scaling + CUDA graphs in Transformers modelling backend (#48010)

Harry Mellor committed 4 months ago

[ROCm] Fix pooling startup workspace lock (#47912)

Andreas Karatzas committed 4 months ago

[docs] Fix the docs build (#48008)

Martin Hickey committed 4 months ago

[Feat] Add runtime monitor for post-warmup TileLang compilation (#46718)

Roberto L. Castro committed 4 months ago

[CI] GSM8K eval integration test for KV offloading (#46893)

Tyler Michael Smith committed 4 months ago

Allow FlashInfer A2A backends for TRTLLM FP8 MoE Modular (#46661)

Thien Tran committed 4 months ago

[ROCm][Bugfix] Fix empty-tensor .max() crash in AITER FA (#47894)

djramic committed 4 months ago

[Perf] Integrate TRTLLM BF16 MoE Modular Kernel (#45182)

Kaihang Jiang committed 4 months ago

[Core] Move MRV1 `late_interaction_runner.py` out of MRV2 subtree (#48014)

Nick Hill committed 4 months ago