Waking up the Model

The inference server is spinning up on Render's free tier. This usually takes 30–60 seconds.

Waiting… 0s

Connecting to server

Loading TensorFlow runtime

Restoring model weights

Running warm-up inference