TinyNarrator / start.sh
cvpfus
Add CPU Spaces deployment with Modal runtimes
41a3f6b
Raw
History Blame Contribute Delete
550 Bytes
#!/bin/sh
# Startup script for Tiny Narrator on Hugging Face Spaces free CPU.
# Model inference endpoints are configured through environment variables.
set -eu
echo "[tiny-narrator] Starting Tiny Narrator app on ${GRADIO_SERVER_NAME:-0.0.0.0}:${PORT:-${GRADIO_SERVER_PORT:-7860}}"
if [ -n "${LLAMA_CPP_BASE_URL:-}" ]; then
echo "[tiny-narrator] Reader brain endpoint configured at ${LLAMA_CPP_BASE_URL}"
else
echo "[tiny-narrator] LLAMA_CPP_BASE_URL is not configured; reader brain will use deterministic fallback."
fi
exec python app.py