mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-07-04 02:00:23 +00:00
e0e912f49b
* llama : add option to override tensor buffers * ggml : fix possible underflow in ggml_nbytes