[Misc] Clean up "swap_space" (#48549)

wangxiyuan committed 4 months ago

[Bugfix] Skip minimax_m3 tool parser tests when Rust extension is absent (#48523)

Matt Woodson committed 4 months ago

[ROCm][MiniMax-M3][Spec Decode] Support speculative decode with AITER sparse PA (#47984)

Tan Pin Siang committed 4 months ago

[Perf][ROCm] Fix GDN KKT warmup regression on RDNA by avoiding fp32 tl.dot (#45000)

nemanjaudovic committed 4 months ago

[ROCm][Kernel] Add HybridW4A16LinearKernel: Triton prefill + HIP skinny decode (#40977)

Matthias Gehre committed 4 months ago

[ROCm][MiniMax-M2] Dispatch fused QK-norm + AllReduce via AITER (#44849)

akii96 committed 4 months ago

up FI fp8 moe topk to 32 (#44462)

Dan Blanaru committed 4 months ago

[1/N] Add dense MHA path for sparse MLA short sequences (#47327)

Matthew Bonanni committed 4 months ago

[Quantization][INC][ARK] Support INT2 XPU WOQ Linear (#47521)

Zhenzhong Xu committed 4 months ago

add pad-aware swiglu limit kernel (#48287)

gnovack committed 4 months ago

[BugFix] Correct OTEL span start time for Dynamo compilation (#40698)

emricksini-h committed 4 months ago

[Frontend] Expose logprob_token_ids on Python OpenAI endpoints (#43463)

langzhao-netizen committed 4 months ago

[Feat] Add Support for BertForMaskedLM to vLLM (#48463)

Snehlata committed 4 months ago

[BugFix] Initialize model_config for Qwen3-VL MoE (#44863)

wenpengw-nv committed 4 months ago

[Core] Preserve Marconi caching with selective hybrid cache retention (#47782)

Nick Hill committed 4 months ago

Added sliding window attention support for qwen-eagle3 architecture (#47568)

shanjiaz committed 4 months ago

Add DCP + Eagle support for Tokenspeed MLA backends (#48180)

Pavani Majety committed 4 months ago

[ROCm][CI] Unblock `AMD: Language Models Test (Extended Pooling)` (#48513)

Micah Williamson committed 4 months ago

[Bugfix][Frontend] Flush engine reasoning parser at engine-reasoning → tool streaming boundary (#47606)

akii96 committed 4 months ago

[Doc] Add DeepseekV32ForCausalLM to supported_models.md (#48293)

Gavin Morris committed 4 months ago