Optimize ci pipeline for on demand restarts

Author: Simon-Count

ci.yml

@@ -6,7 +6,7 @@ prepare:
       then echo 'Install for gpu'
         && GPU_CHOICE=A LAUNCH_AFTER_INSTALL=False ./start_linux.sh; 
       else echo 'Install for cpu'
-        && GPU_CHOICE=N LAUNCH_AFTER_INSTALL=False ./start_linux.sh; 
+        && INSTALL_EXTENSIONS=False GPU_CHOICE=N LAUNCH_AFTER_INSTALL=False ./start_linux.sh; 
       fi"
   - name: "Download model"
     command: "wget -P /home/user/app/models https://huggingface.co/TheBloke/TinyLlama-1.1B-Chat-v0.3-GGUF/resolve/main/tinyllama-1.1b-chat-v0.3.Q6_K.gguf"   
@@ -21,4 +21,9 @@ run:
   # Put your commands here.
   steps:
   - name: "Run"
-    command: "./start_linux.sh --model tinyllama-1.1b-chat-v0.3.Q6_K.gguf"
+    command: "if [ -v NV_LIBCUBLAS_VERSION ];
+      then echo 'Run for gpu'
+        && GPU_CHOICE=A ./start_linux.sh --model tinyllama-1.1b-chat-v0.3.Q6_K.gguf; 
+      else echo 'Run for cpu'
+        && INSTALL_EXTENSIONS=False GPU_CHOICE=N ./start_linux.sh --model tinyllama-1.1b-chat-v0.3.Q6_K.gguf; 
+      fi"