microsoft

/

unilm-base-cased

like 5

Follow

Microsoft 6.22k

Hi,

This model is an extractive BERT-like model, it doesn't generate text. I'd recommend checking out generative models like GPT-Neo, GPT-2, GPT-J, etc.

Hi nielsr,
Thanks for the reply, in their paper https://arxiv.org/pdf/1905.03197.pdf, they have shown that it can be used for Question Generation as well.

Oh cool, didn't know that. In that case you can do the following:

from transformers import AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("microsoft/unilm-base-cased")

and use it like GPT-2 and friends (using the generate method to generate text).

This is the error I am getting by doing above...... :(

I get the same error. How do I interact with this model?

Any update on using this model?