nvjullin b1a63d1b3b

[BugFix] Make FlashInferMetadataBuilder non-blocking (#25040) ...

Signed-off-by: Julien Lin <jullin@nvidia.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>

2025-09-19 20:36:34 +00:00

..

attention

[BugFix] Make FlashInferMetadataBuilder non-blocking (#25040)

2025-09-19 20:36:34 +00:00

core

[bugfix] fix structured outputs key missing issue from #24929 (#25195)

2025-09-19 18:37:57 +00:00

engine

[Frontend] Pass API server count to each process (#23717)

2025-09-20 01:15:19 +08:00

executor

[P/D]kv_output_aggregator support P TP > D TP (#23917)

2025-09-15 11:36:06 +02:00

kv_offload

[KV offload][4/N] Offloading KV connector (#22595)

2025-09-19 19:07:17 +00:00

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

Move ModelConfig from config/__init__.py to config/model.py (#25252)

2025-09-19 16:22:33 +00:00

spec_decode

[Spec Decode] Efficient padded speculation (#24539)

2025-09-18 01:07:24 -04:00

structured_output

[bugfix] fix structured outputs key missing issue from #24929 (#25195)

2025-09-19 18:37:57 +00:00

worker

[Core] Modify the initialization parameters of the lora manager (#25249)

2025-09-19 18:01:28 +00:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

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

cudagraph_dispatcher.py

[CI] Fix linting error on main (#23835)

2025-08-28 06:52:01 -07:00

kv_cache_interface.py

[Core] Get num_encoder_tokens from scheduler config (#24989)

2025-09-16 20:59:07 -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

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

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

utils.py

Allow users to specify kv cache memory size (#21489)

2025-09-11 13:41:07 +00:00