Update app.py

app.py

@@ -10,15 +10,22 @@ import ollama
 # List of available models for selection.
 # IMPORTANT: These names must correspond to models that have been either
 
+ollama pull hf.co/bartowski/Qwen_Qwen3-4B-Instruct-2507-GGUF:Q4_K_M
+#ollama pull hf.co/bartowski/Qwen_Qwen3-4B-Thinking-2507-GGUF:Q4_K_M
+ollama pull smollm2:360m-instruct-q5_K_M
+ollama pull hf.co/bartowski/Llama-3.2-3B-Instruct-GGUF:Q4_K_M
+#ollama pull gemma3n:e2b-it-q4_K_M #slow on Spaces CPU
+ollama pull granite3.3:2b
+ollama pull hf.co/bartowski/tencent_Hunyuan-4B-Instruct-GGUF:Q4_K_M
+
+
 # Model from run.sh
 AVAILABLE_MODELS = [
-    'hf.co/bartowski/Qwen_Qwen3-1.7B-GGUF:Q4_K_M',
     'hf.co/bartowski/Qwen_Qwen3-4B-Instruct-2507-GGUF:Q4_K_M',
     #'hf.co/bartowski/Qwen_Qwen3-4B-Thinking-2507-GGUF:Q4_K_M',
-    'hf.co/bartowski/Llama-3.2-3B-Instruct-GGUF:Q4_K_M', # OK speed with CPU
     'smollm2:360m-instruct-q5_K_M',
-    #'hf.co/unsloth/gemma-3-4b-it-qat-GGUF:Q4_K_M', #Slow.
-    'gemma3n:e2b-it-q4_K_M',
+    'hf.co/bartowski/Llama-3.2-3B-Instruct-GGUF:Q4_K_M', # OK speed with CPU
+    #'gemma3n:e2b-it-q4_K_M',
     'granite3.3:2b',
     'hf.co/bartowski/tencent_Hunyuan-4B-Instruct-GGUF:Q4_K_M'
 ]
@@ -29,12 +36,10 @@ AVAILABLE_MODELS = [
 
 
 # Default System Prompt
-DEFAULT_SYSTEM_PROMPT = """
+DEFAULT_SYSTEM_PROMPT = """Answer everything in simple, smart, relevant and accurate style. No chatty! Besides, pls:
     1. 如果查詢是以中文輸入，使用標準繁體中文回答，符合官方文書規範 
     2. 要提供引用規則依据
-    3. 如果查詢是以英文輸入，使用英文回答
-    4. Answer everything in simple, smart, relevant and accurate style. No chatty!
-    """
+    3. 如果查詢是以英文輸入，使用英文回答"""
 
 # --- Gradio Interface ---
 with gr.Blocks(theme=gr.themes.Default(primary_hue="blue", secondary_hue="neutral")) as demo: