We also use transfer learning a lot. Using transfer learning is but tricky because sometimes you might upset generalised weights of pretrained network with bad

We also use transfer learning a lot. Using transfer learning is but tricky because sometimes you might upset generalised weights of pretrained network with bad hyperparameters. We had written a blog before for using transfer learning as well https://medium.com/nanonets/nanonets-how-to-use-deep-learning-when-you-have-limited-data-f68c0b512cab https://medium.com/nanonets/nanonets-how-to-use-deep-learnin...