Cyrus Leung afffd3cc8a

[Model] Pass mm_features directly into get_mrope_input_positions (#28399) ...

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

2025-11-11 21:14:48 +08:00

..

attention

[Attention] Refactor CUDA attention backend selection logic (#24794)

2025-11-11 07:40:44 -05:00

core

[Misc] Add more scoping for improved trace (#28329)

2025-11-10 21:03:21 +00:00

engine

[Perf] Move gc.freeze logic from EngineCoreProc to EngineCore for better coverage (#27896)

2025-11-10 15:34:18 -08:00

executor

[Core] Simplify async KV output aggregation (#28327)

2025-11-09 09:44:13 -08:00

kv_offload

[KV offload] Enable CPU KV offload on CUDA alike Platforms (#27770)

2025-10-30 22:10:29 +08:00

metrics

[Metrics] Refactor LoRA state tracking (#26801)

2025-11-10 16:34:36 +08:00

pool

…

sample

[CPU]Avoid repeated random sample compile (#28260)

2025-11-07 11:03:57 +00:00

spec_decode

[Attention] Refactor CUDA attention backend selection logic (#24794)

2025-11-11 07:40:44 -05:00

structured_output

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

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

worker

[Model] Pass mm_features directly into get_mrope_input_positions (#28399)

2025-11-11 21:14:48 +08:00

__init__.py

…

cudagraph_dispatcher.py

…

kv_cache_interface.py

[Hybrid] Pass kernel block size to builders (#27753)

2025-11-03 05:48:03 +00:00

outputs.py

Fix cu_num_generated_tokens slicing logic in LogprobsLists.slice() method (#28214)

2025-11-09 19:11:46 +00:00

request.py

[Feature]: Add corrupted request metric to V1 metrics system. (#27306)

2025-11-05 13:45:29 -08:00

serial_utils.py

[Misc] Clean up more utils (#27567)

2025-10-27 15:30:38 +00:00

utils.py

[Misc] Clean up more utils (#27567)

2025-10-27 15:30:38 +00:00