start.sh

#!/bin/bash

# Set Ollama environment variables
export OLLAMA_HOST=0.0.0.0:11434
export OLLAMA_ORIGINS="*"
export OLLAMA_DATA_DIR=/tmp/.ollama
export OLLAMA_MODELS=/tmp/.ollama/models

# Create directories with proper permissions
mkdir -p /tmp/.ollama/models
chmod 755 /tmp/.ollama
chmod 755 /tmp/.ollama/models

# Start Ollama server in background with explicit data directory
echo "Starting Ollama server..."
OLLAMA_DATA_DIR=/tmp/.ollama ollama serve &

# Store the PID
OLLAMA_PID=$!

# Wait for Ollama to start
echo "Waiting for Ollama server to start..."
for i in {1..30}; do
  if curl -s http://localhost:11434/api/version > /dev/null 2>&1; then
    echo "Ollama server started successfully!"
    break
  fi
  echo "Attempt $i/30: Waiting for Ollama..."
  sleep 2
done

# Check if Ollama actually started
if ! curl -s http://localhost:11434/api/version > /dev/null 2>&1; then
  echo "Error: Ollama server failed to start"
  exit 1
fi

# Pull the LLaVA model for vision analysis
echo "Pulling LLaVA model for vision analysis..."
OLLAMA_DATA_DIR=/tmp/.ollama ollama pull llava:7b

if [ $? -eq 0 ]; then
    echo "Model pulled successfully!"
else
    echo "Warning: Failed to pull model, but continuing..."
fi

# Start FastAPI on port 7860 (HF Spaces requirement)
echo "Starting FastAPI server on port 7860..."
cd /app
python3 -m uvicorn fast:app --host 0.0.0.0 --port 7860