Closed taewookim closed 6 years ago
@taewookim I've tried to run it on cpu, but it does not work, because model was trained in gpu.
really? is that some kind of pytorch limitation? pretty sure in TF, you can train on GPU and run on CPU no problem...
@taewookim I tried with no luck, of course you are welcome to pull request with your implementation.
thanks @smivv . will poke around and see if i find anything
@taewookim let me know if you need something.
what modifications would i need to make?