Similarity models - improve

Use a pretrained model
If you are trying to solve a similarity problem, you can use a pretrained block
When you use pretrained blocks you do transfer learning. Transfer learning is when you train a model on a large dataset (such as ImageNet for image data) and want to transfer parts of that knowledge to solve another, related problem.
If a model is already pretrained for the purpose of similarity you don’t need to train it. You can click Run and wait until the experiment has finished the default 1 epoch.

Check your data quality, if you are not getting the desired results.