Woosuk Kwon b9926f7f66 Support block size 32 (#35)

2023-04-09 23:07:18 -07:00

..

http_frontend

Add an option to use dummy model weights (#33)

2023-04-08 23:36:12 -07:00

master

Support block size 32 (#35)

2023-04-09 23:07:18 -07:00

models

Add an option to use dummy model weights (#33)

2023-04-08 23:36:12 -07:00

parallel_utils

Add CUDA graph-based all reduce launcher (#26)

2023-04-05 11:16:57 -07:00

worker

Add an option to use dummy model weights (#33)

2023-04-08 23:36:12 -07:00

block.py

Support beam search & parallel generation (#7)

2023-03-10 09:58:21 -08:00

sampling_params.py

FastAPI-based working frontend (#10)

2023-03-29 14:48:56 +08:00

sequence.py

Implement preemption via recomputation & Refactor scheduling logic (#12)

2023-03-30 14:51:46 -07:00

utils.py

FastAPI-based working frontend (#10)

2023-03-29 14:48:56 +08:00