Forward Anthropic vllm_xargs to sampling params (#53308)

Maroon Ayoub committed 2 months ago

Fix Cohere ChatV2 citation and tool handling issues (#52175)

andrewbcohere committed 2 months ago

[CI/Build][ROCm] Run the TileLang HIP symbol checks in their own interpreter (#53117)

stefankoncarevic committed 2 months ago

Revert "Remove native Hunyuan V1 and VL implementations" (#53296)

Kevin H. Luu committed 2 months ago

Revert "[ROCm][Perf] Kimi-K3 Fused kernels for KDA prefill" (#53294)

Kevin H. Luu committed 2 months ago

[Bugfix][KV Connector][NIXL] Support PCP producers (#52779)

Lucas Wilkinson committed 2 months ago

[CI] Fix MultiConnector accuracy test lifecycle (#53023)

Andreas Karatzas committed 2 months ago

[ROCm][CI] aiter kernel ops - enable rope test (#52854)

Divakar Verma committed 2 months ago

[ROCm][CI] Stabilize MI355 FlyDSL MoE accuracy test (#53024)

Andreas Karatzas committed 2 months ago

[ROCm][CI] Add float16 dtype and unsupported head size tests for paged attention (#53177)

Divakar Verma committed 2 months ago

[ROCm][CI] Stabilize MI355 FusedMoE test group (#53025)

Andreas Karatzas committed 2 months ago

Revert compile-cache device index regression on CPU (#53304)

Kevin H. Luu committed 2 months ago

[ROCm][Test] Use platform FP8 dtype in ModelOpt FP8_PB_WO test (#53268)

djramic committed 2 months ago

[DSv4 Perf] Adaptive topk width for dsv4, making #50004 back (#52823)

Wentao Ye committed 2 months ago

[Cleanup][MLA] Remove FlashInfer DSpark DCP support (#53139)

Summer Yang committed 2 months ago

[Bugfix][Attention] Fall back to native FlashInfer decode when XQA cannot serve a KV-cache group's head_dim (#53111)

Stefano Castagnetta committed 2 months ago

Revert "[Bugfix][MoE] Tune FlashInfer experts to scheduler token limit" (#52989) (#53186)

vllm-agent committed 2 months ago

[K3 Perf] Fuse MXFP4 top-k finalization into latent-tail, ~5% E2E latency reduction (#53152)

Wentao Ye committed 2 months ago

[ROCm][Perf] Optimize DeepSeek V4 C4A top-k with AITER (#52882)

Fangzhou Ai committed 2 months ago

[Frontend] Use VLLMValidationError for batch request URL validation (#50191)

Tan Chao committed 2 months ago