Electra support

[samanjoy2]

Is there any way to use an electra bert model with this?

[crux82]

You should change the model specified by the variable model_name

[samanjoy2]

Using model_name = 'google/electra-base-discriminator'

Gives me this error, `======== Epoch 1 / 10 ======== Training...

RuntimeError Traceback (most recent call last) in 78 # Generate the output of the Discriminator for real and fake data. 79 # First, we put together the output of the tranformer and the generator ---> 80 disciminator_input = torch.cat([hidden_states, gen_rep], dim=0) 81 # Then, we select the output of the disciminator 82 features, logits, probs = discriminator(disciminator_input)

RuntimeError: Tensors must have same number of dimensions: got 2 and 3`