mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-10-31 08:51:55 +00:00
1da7b76569
* server : fix speculative decoding with context shift ggml-ci * server : take into account speculative limits ggml-ci * server : add tests
3.8 KiB
3.8 KiB