Lianmin ZhengandSangBin Cho rkooo567@gmail.com 8a6906127a Improve linear.py to load sharded weights & remove the dependency of Parameters from vllm (#2784) ...

Co-authored-by: SangBin Cho rkooo567@gmail.com

2025-01-07 23:29:10 -08:00

..

data_parallel_controller.py

Misc fix for min_p_sampling, --cuda-graph-bs (#2761)

2025-01-07 02:52:53 -08:00

detokenizer_manager.py

[Feature] Get Token IDs with Engine.generate() (#2636)

2024-12-29 12:28:27 -08:00

image_processor.py

Simplify tokenizer manager (#2254)

2024-11-29 02:18:51 -08:00

io_struct.py

Speed up update_weights_from_tensor (#2695)

2025-01-02 02:05:19 -08:00

schedule_batch.py

Clean up eagle code (#2756)

2025-01-06 14:54:18 -08:00

schedule_policy.py

Refactor SchedulePolicy to improve code organization (#2571)

2025-01-04 00:05:16 +08:00

scheduler.py

Misc fix for min_p_sampling, --cuda-graph-bs (#2761)

2025-01-07 02:52:53 -08:00

session_controller.py

Improve linear.py to load sharded weights & remove the dependency of Parameters from vllm (#2784)

2025-01-07 23:29:10 -08:00

tokenizer_manager.py

Fix package loss for small models (#2717)

2025-01-02 18:25:26 -08:00

tp_worker_overlap_thread.py

Refactor logprob computation to return the real logprob used in sampling (#2664)

2024-12-30 04:51:38 -08:00

tp_worker.py

Eagle speculative decoding part 3: small modifications to the general scheduler (#2709)

2025-01-02 02:09:08 -08:00