Update run.sh

run.sh

@@ -4,23 +4,18 @@
 ollama serve &
 
 # Wait a few seconds for the server to be fully operational
-sleep 5
+sleep 10
 
 # Pull the model from Hugging Face.
 # This command downloads the model and makes it available to the API.
 # It's the non-interactive version of "ollama run".
-ollama pull hf.co/bartowski/Qwen_Qwen3-4B-Instruct-2507-GGUF:Q4_K_M
+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 '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
-
-
-#---- 20250722 fail to run
-#ollama pull hf.co/ggml-org/SmolLM3-3B-GGUF:Q4_K_M
-#ollama pull hf.co/bartowski/nvidia_OpenReasoning-Nemotron-1.5B-GGUF:Q5_K_M
+ollama pull 'granite3.3:2b'
+ollama pull 'hf.co/bartowski/tencent_Hunyuan-4B-Instruct-GGUF:Q4_K_M'
 
 
 # Start the Gradio web application