Lucas Wilkinson dc464a3d39 [BugFix] AssertionError: Do not capture num_reqs > max_num_reqs for uniform batch (#25505) ...

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

2025-09-23 18:00:29 -06:00

..

attention

[V1][Kernel] Add triton implementation for reshape_and_cache_flash (#24503)

2025-09-23 12:52:40 -04:00

core

[Core] Use KVCacheBlock as much as possible instead of dict[block_id, KVCacheBlock] (#24830)

2025-09-23 15:11:14 -07:00

engine

[DP] support torchrun external launcher with Data Parallelism (#24899)

2025-09-22 12:06:05 -07:00

executor

[V0 Deprecation] Remove V0 MP executor (#25329)

2025-09-20 21:26:35 -07:00

kv_offload

[KV offload][5/N] Add CPUOffloadingSpec (#24251)

2025-09-22 12:30:36 -07:00

metrics

[P/D][Nixl] Introduce KVTransferMetrics and aggregation strategy (#22188)

2025-09-19 11:09:14 +00:00

pool

[Performance] V1 Pooling Models E2E Performance Optimization (#23162)

2025-08-21 13:26:09 +00:00

sample

[Docs] Fix warnings in mkdocs build (continued) (#25042)

2025-09-20 11:45:18 +00:00

spec_decode

[Speculators][Speculative Decoding] Fix gpt-oss eagle3 accuracy issue (#25406)

2025-09-23 15:44:35 -04:00

structured_output

[Build] Update Xgrammar to 0.1.25 (#25467)

2025-09-23 21:25:46 +00:00

worker

[BugFix] AssertionError: Do not capture num_reqs > max_num_reqs for uniform batch (#25505)

2025-09-23 18:00:29 -06:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

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

cudagraph_dispatcher.py

[torch.compile] CUDAGraph Inductor partition integration (#24281)

2025-09-20 01:02:15 +00:00

kv_cache_interface.py

[Hybrid Allocator] Support full attention with different hidden size (#25101)

2025-09-19 23:43:59 -07:00

outputs.py

[P/D][Nixl] Introduce KVTransferMetrics and aggregation strategy (#22188)

2025-09-19 11:09:14 +00:00

request.py

[CORE] Prompt Embeddings Support for v1 Engine (#24278)

2025-09-19 08:03:09 +08:00

serial_utils.py

[Misc] Support more collective_rpc return types (#25294)

2025-09-20 02:02:38 +00:00

utils.py

Add VLLM_NVTX_SCOPES_FOR_PROFILING=1 to enable nvtx.annotate scopes (#25501)

2025-09-23 22:50:09 +00:00