Load and extract the model for language modeling

#123
by theodp - opened

Hello,

Currently looking for a way to fine-tune the model on a modeling task.

I locally downloaded files from this repository.

I've been able to initialize the model locally with AutoModelForCausalLM.from_pretrained() function.

I can use it for generation, but I'd like to know how can I extract the "Transformer class" from the "MistralModel" object in order to use it for next token prediction please ?

Thank you !

You can use AutoModel.from_pretrained()

Sign up or log in to comment