[ROCm] Enable gluon paged MQA logits on gfx950 (MI355X) (#42062)

frida-andersson committed 6 months ago

[V1][DP][LB] Publish request counts at the start of each engine step (#41626)

Vadim Gimpelson committed 6 months ago

PD disagg with NIXL Connector: GDN support (Qwen3.5) (#41869)

zhanqiuhu committed 6 months ago

[RFC] Replace shared-memory routed experts with ModelRunnerOutput transfer and HTTP support (#39568)

Hongxin Xu committed 6 months ago

[Bugfix] Fix TRTLLM ragged MLA prefill workspace warmup (#42112)

Mohammad Miadh Angkad committed 6 months ago

[kv_offload] Implement `reset_cache()` for the offloading connector (#41956)

Martin Hickey committed 6 months ago

[Compile] Fix compile warning with topk softplus sqrt (#41261)

Wentao Ye committed 6 months ago

[DSV4] Fuse norm and router for low latency scenario (#41263)

Jee Jee Li committed 6 months ago

[Quantization][Autoround][Toolkit] Add W4A16 Support (#39778)

Zhenzhong Xu committed 6 months ago

[XPU] Fix double-transpose in XPUFP8ScaledMMLinearKernel for W8A8 quant method (#41689)

Libin Tang committed 6 months ago

[Bug] Fix DeepSeek V4 `AttributeError: module 'cutlass.cute.nvgpu' has no attribute 'LoadCacheMode'` (#42342)

Wentao Ye committed 6 months ago

[CI][AMD][BugFix] Prevent triton compiler error when running test_moe_layer with use_ep = True on ROCm (#40857)

rasmith committed 6 months ago

[CI][AMD] Skip tests where models have problems or fails on both HW types (#42126)

rasmith committed 6 months ago

Update Dockerfile.rocm for AINIC & Thor NIC (#40453)

haic0 committed 6 months ago

Revert "[Core] Replace routing replay with device cache and async D2H pipeline" (#39917) (#42434)

aoshen02 committed 6 months ago

[MLA Attention Backend] Add TOKENSPEED_MLA backend for DSR1/Kimi K25 prefill + decode on Blackwell (#41778)

Yongye Zhu committed 6 months ago

Use hidden_pad and intermediate_pad from vLLM #34301 (#42098)

Rebecca Lee committed 6 months ago

[CI][XPU] skip ut of offload connector (#42598)

liuzhenwei committed 6 months ago

[Misc] Fix mypy error in parser_manager type narrowing (#42441)

Sarah Salah committed 6 months ago

[Fix] Weight loading for qwen3_5 using runai_streamer (#42521)

Harsh Shah committed 6 months ago