add --no-mmap in llama-bench (#5257)

* add --no-mmap, show sycl backend * fix conflict * fix code format, change print for --no-mmap * ren no_mmap to mmap, show mmap when not default value in printer * update guide for mmap * mv position to reduce model reload
2024-02-02 03:48:53 +08:00
parent 4d0924a890
commit 128dcbd3c9
4 changed files with 89 additions and 10 deletions
@@ -405,7 +405,7 @@ Using device **0** (Intel(R) Arc(TM) A770 Graphics) as main device

  llama.cpp use mmap as default way to read model file and copy to GPU. In some system, memcpy will be abnormal and block.

-  Solution: add **--no-mmap**.
+  Solution: add **--no-mmap** or **--mmap 0**.

 ## Q&A