[CPU][CI] Enable fused MoE tests in Arm CI (#30132)

Fadi Arafeh committed 11 months ago

update torchao safetensors impl (#30155)

liangel-02 committed 11 months ago

kv_transfer: Rename the shared storage connectors (#30201)

Or Ozeri committed 11 months ago

[Bugfix] Fix compressed-tensors models failing to load with transformers backend (#30287)

Michael Goin committed 11 months ago

[Kernel]Support W4A8 Grouped GEMM on Hopper (#29691)

czhu-cohere committed 11 months ago

Lora MoE Align Improvements (#29257)

gnovack committed 11 months ago

[ci] Refactor CI file structure (#29343)

Kevin H. Luu committed 11 months ago

[ROCm][CI] Skip NVIDIA-Only Prime-RL Test in AMD CI (#29420)

Micah Williamson committed 11 months ago

[Bugfix] Skip generation config fallback for GGUF to prevent multi-process hang (#30209)

Christina Norman committed 11 months ago

Mark qwen2_5_vl as xfail (#30283)

Yanan Cao committed 11 months ago

[Bugfix] Fix DeepGEMM after #29546 (#30267)

Zhewen Li committed 11 months ago

[Perf] Optimize `group_topk` kernel, 1.9% Throughput improvement, 2.1% TPOT improvemnt (#30159)

Wentao Ye committed 11 months ago

[Feature] Batch invariant: Enable `TRITON_MLA` without prefix-caching (#29125)

Wentao Ye committed 11 months ago

[moe] Allow disabling DP chunking (#29936)

Ming Yang committed 11 months ago

feat(metrics): Add prefill KV compute metric excluding cached tokens (#30189)

Victor Ziliang Peng committed 11 months ago

[Disagg] Support large batch size in proxy server and update NixlConnector doc for DP (#28782)

Ming Yang committed 11 months ago

[Perf] Improve fp8 quant in mla; replace ReduceSum with ReduceScatterSum (#29795)

Lain committed 11 months ago

[ROCm][CI] Fix test_max_len.py for Rocm (#29916)

Charlie Fu committed 11 months ago

Add SpecDec support to `selective_state_update` (#29488)

roikoren755 committed 11 months ago

Bump nvshmem to 3.3.24 and fix CUDA 13 installation (#30149)

Dmitry Tokarev committed 11 months ago