Martin Hickey f4417f8449

[KVConnector] Add KV events to KV Connectors (#28309) ...

Signed-off-by: Martin Hickey <martin.hickey@ie.ibm.com>

2025-12-11 15:30:29 +01:00

..

attention

[perf] Use direct copy (broadcast) instead of cat for k_nope/k_pe in MLA prefill (#29710)

2025-12-11 08:20:45 +00:00

core

[Bugfix] fix confusing OOM errors during v1 init (#28051)

2025-12-10 23:17:41 +00:00

engine

[Perf] Enable environment cache in EngineCore to enable the feature for UniProcExecutor as well (#29289)

2025-12-10 14:13:01 -05:00

executor

…

kv_offload

[Attention] Update attention imports (#29540)

2025-11-27 11:19:09 -05:00

metrics

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

2025-12-09 00:08:48 +00:00

pool

[Model][6/N] Improve all pooling task | Support chunked prefill with ALL pooling (#27145)

2025-12-04 13:44:15 +00:00

sample

[Perf] Avoid pageable HtoD transfer in MinTokensLogitsProcessor (#29826)

2025-12-02 21:25:52 +00:00

spec_decode

[Attention] Make seq_lens_cpu optional in CommonAttentionMetadata to enable true async spec-decode (#29624)

2025-12-09 17:18:10 -08:00

structured_output

[Bugfix] Fix the issue where DeepSeek v3.2 cannot use structured_output (#30371)

2025-12-10 08:30:16 +00:00

worker

[KVConnector] Add KV events to KV Connectors (#28309)

2025-12-11 15:30:29 +01:00

__init__.py

…

cudagraph_dispatcher.py

[BugFix] Fix assert batch_descriptor.num_tokens == num_tokens_padded (#30173)

2025-12-09 10:36:12 -05:00

kv_cache_interface.py

…

outputs.py

[KVConnector] Add KV events to KV Connectors (#28309)

2025-12-11 15:30:29 +01:00

request.py

[P/D] KV Load Failure Recovery/Abort Configuration (#26813)

2025-12-10 11:00:52 -08:00

serial_utils.py

[Model] Move multimodal_cpu_fields definition to field config (#30181)

2025-12-06 13:40:02 +00:00

utils.py

…