RuntimeError: Expected object of scalar type Long but got scalar type Int for argument #2 ‘target’ in call to _thnn_nll_loss_forward

fastai fastai dev
1

I am running into this error with fastai2, originally posted here and was fixed Windows: RuntimeError: cuda runtime error (801) & RuntimeError: Expected object of scalar type Long but got scalar type

Using Windows

Fastai Version: 0.0.5
Pytorch Version: 1.3.1
Python version: 3.7.6 (default, Jan 8 2020, 20:23:39) [MSC v.1916 64 bit (AMD64)]

Error:

functional_
functional_.PNG1139×699 48.1 KB

Fixed the issue by using

target = target.long()

Wondering if the data processing pipeline is not automatically converting to long ints (like last time) or an issue on my side.

Thanks

2

I used target = target.long(). I am still getting the same error. Please help @amritv .

3

Hey @mohitasudani, what dataset are you using and what verison of fastai2. Also pls include the error trail