mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-02 09:12:03 +00:00
abd4d0bc4f
* speculative : update default params * speculative : do not discard the last drafted token
29 KiB
29 KiB