Create start.sh

start.sh

@@ -0,0 +1,53 @@
+#!/bin/bash
+set -e
+
+echo "🚀 Starting Ollama + FastAPI Server..."
+
+# Set Ollama environment variables
+export OLLAMA_HOST=0.0.0.0:11434
+export OLLAMA_ORIGINS="*"
+export OLLAMA_KEEP_ALIVE=5m
+
+# Start Ollama in background (as root - this is fine in containers)
+echo "📡 Starting Ollama service..."
+ollama serve &
+OLLAMA_PID=$!
+
+# Wait for Ollama to be ready
+echo "⏳ Waiting for Ollama to be ready..."
+for i in {1..60}; do  # Increased timeout to 60 attempts
+    if curl -f http://127.0.0.1:11434/api/tags >/dev/null 2>&1; then
+        echo "✅ Ollama is ready!"
+        break
+    fi
+    echo "🔄 Attempt $i/60 - Waiting for Ollama..."
+    sleep 2
+done
+
+# Check if Ollama is actually running
+if ! curl -f http://127.0.0.1:11434/api/tags >/dev/null 2>&1; then
+    echo "❌ Ollama failed to start properly"
+    echo "🔍 Checking Ollama process..."
+    ps aux | grep ollama || echo "No Ollama process found"
+    
+    echo "🔄 Attempting to restart Ollama..."
+    killall ollama 2>/dev/null || true
+    sleep 2
+    ollama serve &
+    sleep 10
+fi
+
+# Try to pull the model (no user switching)
+echo "📥 Pulling Llama 3.2 1B model..."
+ollama pull hf.co/bartowski/Llama-3.2-1B-Instruct-GGUF:Q4_K_M || {
+    echo "⚠️ Model pull failed, trying smaller model..."
+    ollama pull llama3.2:1b || echo "⚠️ Fallback model pull also failed, continuing..."
+}
+
+echo "🎉 Setup complete!"
+echo "📡 Ollama API: http://localhost:11434"
+echo "📖 FastAPI: http://localhost:7860"
+
+# Start FastAPI in foreground (keeps container alive)
+echo "🚀 Starting FastAPI server..."
+python app.py