scripts.models.eval.eval

Perform inference with a trained TERMinator model.

The resulting evaluated proteins will be dumped in <output_dir> via a pickle file net.out.

Usage:

python eval.py \
    --dataset <dataset_dir> \
    --model_dir <trained_model_dir> \
    --output_dir <output_dir> \
    [--subset <data_subset_file>] \
    [--dev <device>]

If subset is not provided, the entire dataset dataset will be evaluated.

See python eval.py --help for more info.