Update entrypoint.sh

entrypoint.sh

@@ -1,19 +1,24 @@
 #!/bin/bash
-set -e
 
-echo "Starting Ollama..."
+# 1. Start Ollama in the background
+echo "Starting Ollama server..."
 ollama serve &
 
-# Wait for Ollama to wake up
-sleep 10
+# 2. Wait for Ollama to wake up (Checking every second)
+echo "Waiting for Ollama to be ready..."
+while ! curl -s http://localhost:11434/api/tags > /dev/null; do
+    sleep 1
+done
+echo "✅ Ollama is ready!"
 
-echo "Pulling standard models..."
-ollama pull qwen2.5-coder:1.5b
-ollama pull qwen2.5-coder:3b
-ollama pull qwen2.5-coder:7b
+# 3. Download the specific 30B model
+echo "📥 Downloading Qwen 30B... (This will take a while, check logs)"
+ollama pull hf.co/bartowski/Qwen_Qwen3-30B-A3B-GGUF:Q4_K_M
 
-# NOTE: The 30B model download is now handled safely inside app.py
-# We removed the manual 'huggingface_hub' download block to prevent startup crashes.
+# 4. Download the smaller fast models (optional, but good to have)
+echo "📥 Downloading smaller models..."
+ollama pull qwen2.5-coder:3b
 
-echo "Starting app..."
+# 5. Finally, start the Python App
+echo "🚀 Starting App..."
 python3 /app.py