You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
I would like to ask a question about generative pretraining using GPT. As far as I know, GPT uses the transformer decoder as well and one of the main train scheme is to predict next tokens. In scGPT pretraining, what "next token"corresponds to? I did not get the difference between BERT architecture of scGPT architecture?
The text was updated successfully, but these errors were encountered:
Hi,
I would like to ask a question about generative pretraining using GPT. As far as I know, GPT uses the transformer decoder as well and one of the main train scheme is to predict next tokens. In scGPT pretraining, what "next token"corresponds to? I did not get the difference between BERT architecture of scGPT architecture?
The text was updated successfully, but these errors were encountered: