resolve GPU tokenize padding issue

[WeihaoGe1009]

Currently, when running the code on a GPU cluster, it doesn't recognize padding.

I got an error of:

Asking to pad, but the tokenizer does not have a padding token. 
Please select a token to use as 'pad_token'
'(tokenizer.pad_token = tokenizer.eos_token e.g.)' 
or add a new pad token via 'tokenizer.add_special_tokens({'pad_token': '[PAD]'})'.

This issue doesn't occur when I was running on the CPU-only cluster.

On GPU the following codes got ignored:

tokenizer.pad_token = tokenizer.eos_token

or

if tokenizer.pad_token is None:
    tokenizer.add_special_tokens({'pad_token': '[PAD]'})
    model.resize_token_embeddings(len(tokenizer))

[WeihaoGe1009]

try without padding