Given your pretrained model, how can I evaluate it on IU dataset?

[pierremaged]

No description provided.

[zaaachos]

@pierremaged First of all, you should comment out the code-block on the red rectangles in trainer.py. This will make the model run on evaluation mode

Secondly, after downloading the pre-trained model, copy the path location of the downloaded model file, and type in the --resume argument a default value with the path location in string.
something like this:
parser.add_argument('--resume', default='path\to\model\file', type=str, help='whether to resume the training from existing checkpoints.')
You could find this in main.py file.