[TPU] Support Pathways in vLLM (#21417)

wenxindongwork committed over 1 year ago

feat(distributed): add `get_required_kvcache_layout` class method to kv connector api (#20433)

wxsm committed over 1 year ago

[Bugfix] SharedStorage Connector for V1 PD multimodal (#21611)

Chenguang Zheng committed over 1 year ago

[CI/Build] Fix registry tests (#21934)

Cyrus Leung committed over 1 year ago

[Bugfix] Fix TypeError in scheduler when comparing mixed request_id types (#21816)

633WHU committed over 1 year ago

Override attention metadata for fast prefill in some KV sharing setups (#21590)

Yong Hoon Shin committed over 1 year ago

[Bugfix] Fix multi-api server not working for text models (#21933)

Cyrus Leung committed over 1 year ago

[Bugfix] Fix OOM tests in initialization test (#21921)

Isotr0py committed over 1 year ago

[Docs] Reduce the size of the built docs (#21920)

Harry Mellor committed over 1 year ago

[NVIDIA] Fix Llama4 Scout FP4 functionality issues (#21499)

Po-Han Huang (NVIDIA) committed over 1 year ago

[Misc] Improve code readability of KVCacheManager (#21673)

Ruixiang Tan committed over 1 year ago

Reduce time wasted in GitHub Actions using `concurrency` (#21919)

Harry Mellor committed over 1 year ago

[CI/Build] Only run markdownlint in CI (#21892)

Cyrus Leung committed over 1 year ago

[Test] Add Benchmark and Unit Test for `per_token_group_quant` (#21860)

Wentao Ye committed over 1 year ago

[misc] skip p2p check by default (#21904)

youkaichao committed over 1 year ago

[Feature] Support multiple api keys in server (#18548)

Yan Pashkovsky committed over 1 year ago

[Bugfix] we should use metavar is not choices (#21902)

rongfu.leng committed over 1 year ago

[Doc] Remove vLLM prefix and add citation for PagedAttention (#21910)

Cyrus Leung committed over 1 year ago

[Feature][EPLB] Add eplb support for Qwen3 (#20815)

aladerran committed over 1 year ago

[Hardware][CPU] Build fix for ARM without BF16 (#21848)

Eric Curtin committed over 1 year ago