[Rust Frontend] Support `--generation-config vllm` (#53044)

Bugen Zhao committed 2 months ago

[CI][Docker] Pin remaining manylinux builder images (#53172)

Kevin H. Luu committed 2 months ago

[Frontend][Core][Spec Decode] Per-request acceptance stats in OpenAI API responses (#48915)

Matthew Kotila committed 2 months ago

Fix seed loss when batch contains unseeded requests (#51866)

MKQuantum committed 2 months ago

[CI][Docker] Pin manylinux2_28-builder:cuda13.0 to the release/2.13 image (#52994)

Andrey Talman committed 2 months ago

[Bugfix][MoE] Tune FlashInfer experts to scheduler token limit (#52989)

Misha Goin committed 2 months ago

[DSV4][Kernel] Fuse shared experts into MegaMoE (#53040)

Canlin Guo committed 2 months ago

[CI/Build][ROCm] Keep the CUDA-only kernel tests out of the ROCm run (#53113)

stefankoncarevic committed 2 months ago

[Bugfix] Support MistralCommonBackend tokenizers in structured output (#52720)

Thanh Phan committed 2 months ago

[Bugfix][MiniMax-M3] Keep FP8 query allocation stable across CUDA graph replay (#51203)

kyleliang-nv committed 2 months ago

[Misc] Don't allow language-model-only used with encoder CG together (#53127)

Isotr0py committed 2 months ago

upgrade tpu-inference to v0.27.0 (#53088)

Dennis Yeh committed 2 months ago

Reduce `AutoWeightsLoader` kwargs (#53106)

Harry Mellor committed 2 months ago

Fix Transformers modelling backend `RMSNormFuser.fuse` performance (#52766)

Harry Mellor committed 2 months ago

[Feature][Model Runner V2] Support extract_hidden_states speculation (#49811)

Zupeng Wang committed 2 months ago

[Core][Multimodal] Skip redundant placeholder scan when token match succeeds (#52925)

Yiqin committed 2 months ago

[Kernel] Add FlashInfer TRTLLM MXFP8 linear backend (#52204)

Seonjin committed 2 months ago

[Bugfix] Skip MM processor cache inserts larger than capacity (#53016)

Prudhvi Vuda committed 2 months ago

[Docker] Update to nixl-1.3.2 (#51777)

Sandeep Maddipatla committed 2 months ago

[Model] Add tower and connector LoRA support for LFM2-VL (#51498)

Zupeng Wang committed 2 months ago