Transfer Learning

Hello,

Can we use this transfer learning in the real world applications, 'Cause many of them are using TensorFlow neural nets to build real-world applications, for instance, I’ll post some of the links who where built there model using TensorFlow neural nets,

  1. https://youtu.be/uGYJuOyIvzs

  2. https://youtu.be/9r2VVM4nfk8

Please, tell anyone about this.

Thanks & Regards
P.Venkatesh

Hello, Is there anyone to answer?

@bestfeatures
Sorry, no one could reply to your question!
Yes, transfer learning, can, and should be used, wherever possible, even in real world applications. And it is possible to transfer model weights built in TF to Pytorch and vice versa. All you need to do is find a way to change TF tensors to Pytorch Tensors. Model Architectures would be same, i guess! And then, importing weights is not that big of a deal!.
I haven’t tried it ever myself, but there are great guides out there. I am attaching a link that might give you a direction. From there, you can search for what you feel is important and helpful.