Yifan QiaoandChen Zhang 48ddb02b79 [Hybrid Allocator] Support KV cache groups with different block_size (#29143) ...

Signed-off-by: Yifan Qiao <yifanqiao@berkeley.edu>
Co-authored-by: Chen Zhang <zhangch99@outlook.com>

2025-11-25 10:30:57 -05:00

..

attention

[Core] Generalize Encoder-Decoder seq_lens computation to avoid Whisper hardcoded logic (#29268)

2025-11-25 11:32:11 +00:00

core

[Hybrid Allocator] Support KV cache groups with different block_size (#29143)

2025-11-25 10:30:57 -05:00

engine

[BugFix] bad_words filtering ineffective when n > 1 (#29313)

2025-11-25 09:36:34 +00:00

executor

[Feature] Prefill Context Parallel (PCP) basic support (#28718)

2025-11-19 15:52:44 -05:00

kv_offload

[KVConnector][Core] Support cross-layer KV blocks (#27743)

2025-11-20 19:09:59 +01:00

metrics

[Metrics] Scheduled removal of deprecated metrics (#29330)

2025-11-25 11:15:13 +08:00

pool

[Chore] Separate out vllm.utils.platform_utils.py (#27374)

2025-10-23 19:08:06 +00:00

sample

[BugFix] Fix returned logprobs with spec decode + prefill chunking (#29216)

2025-11-22 22:41:25 +08:00

spec_decode

Revert back to torch.equal over torch.allclose from #28819 (#29086)

2025-11-25 14:23:38 +00:00

structured_output

[Bugfix] Fix llguidance backend, rollback when EOS was encountered (#25905)

2025-11-10 14:53:32 -05:00

worker

[Bugfix] Fix overallocation in MM profiling (#29386)

2025-11-25 12:38:36 +00:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

2024-11-11 23:05:38 +00:00

cudagraph_dispatcher.py

[LoRA] LoRA cuda graph specialization (#25914)

2025-10-20 04:21:09 +00:00

kv_cache_interface.py

[Feature] Prefill Context Parallel (PCP) basic support (#28718)

2025-11-19 15:52:44 -05:00

outputs.py

Revert "[Redo] #26368 (#28771)" (#29121)

2025-11-20 21:27:45 -08:00

request.py

[KV Connector] Fix async connector prefix cache metrics (#28585)

2025-11-21 17:45:00 -05:00

serial_utils.py

[DisaggEverything] Tokens in<>out /generate endpoint (#24261)

2025-11-14 09:58:01 -07:00

utils.py

[BugFix] Fix async scheduling + chunked prefill + preemption (#28787)

2025-11-17 06:46:46 +08:00