fix: Vicuna template instead of ChatML

app.py

@@ -50,15 +50,14 @@ def start_ollama():
     # Step 3: Generate Modelfile dynamically
     modelfile_content = f'''FROM {MODEL_FILE}
 
-TEMPLATE """<|im_start|>system
-{{{{ .System }}}}<|im_end|>
-{{{{ range .Messages }}}}<|im_start|>{{{{ .Role }}}}
-{{{{ .Content }}}}<|im_end|>
-{{{{ end }}}}<|im_start|>assistant
+TEMPLATE """{{{{ if .System }}}}{{{{ .System }}}}
+{{{{ end }}}}{{{{ range .Messages }}}}{{{{ if eq .Role "user" }}}}
+USER: {{{{ .Content }}}}
+ASSISTANT: {{{{ end }}}}
+{{{{ end }}}}
 """
 
-PARAMETER stop "<|im_end|>"
-PARAMETER stop "<|im_start|>"
+PARAMETER stop "</s>"
 PARAMETER num_ctx 2048
 '''
     mf_path = "/tmp/Modelfile"