mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-01 09:01:57 +00:00
* retrieval * Reuse querybatch to reduce frequent memory allocation * delete unused white space
10 KiB
10 KiB