openblas

app.py

@@ -1,3 +1,9 @@
+import os
+
+# we need to compile a OPENBLAS version  for cpu
+# Or get it from  https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/
+os.system('CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python==0.2.11')
+
 import re, requests, json
 import gradio as gr
 import random
@@ -6,7 +12,6 @@ from itertools import chain
 import asyncio
 from llama_cpp import Llama
 import datetime
-import os 
 
 from transformers import (
     StoppingCriteriaList,