Inference

Applies to: python, general

Inference is running a trained model to produce predictions, without updating its parameters.

with torch.no_grad():
    y = model(x)

See also: neural-network, estimator