entrypoint.sh

#!/bin/bash
# Customs Compass entrypoint — starts Ollama in background, waits for it
# to be ready, ensures the model is available, then launches Streamlit.

set -e

echo "==> Starting Ollama server in background…"
ollama serve > /tmp/ollama.log 2>&1 &
OLLAMA_PID=$!

# Wait up to 60s for Ollama to accept connections
echo "==> Waiting for Ollama to be ready…"
for i in {1..60}; do
    if curl -sf http://localhost:11434/api/tags > /dev/null 2>&1; then
        echo "==> Ollama is ready (after ${i}s)."
        break
    fi
    sleep 1
done

# Verify the model is present (it was pulled at build time, but double-check)
if ! ollama list | grep -q "llama3.2:1b"; then
    echo "==> Pulling llama3.2:1b (this may take 1-2 minutes)…"
    ollama pull llama3.2:1b
fi

echo "==> Pre-warming the model so first user request is fast…"
curl -sf -X POST http://localhost:11434/api/generate \
    -H "Content-Type: application/json" \
    -d '{"model":"llama3.2:1b","prompt":"hi","stream":false,"keep_alive":"15m"}' \
    > /dev/null 2>&1 || true

echo "==> Launching Streamlit on port ${PORT:-7860}…"
exec streamlit run app.py \
    --server.port="${PORT:-7860}" \
    --server.address=0.0.0.0 \
    --server.headless=true \
    --browser.gatherUsageStats=false