Model fine-tuning and inference of Bloom 560M

🤗Transformers
1

Hi all,

I’m the Co-founder of inferencetraining.ai and openagi.ai

How do I simply fine-tune Bloom 560M and make inferences, post? I’ve followed the steps Finetune BLOOM (Token Classification) | Kaggle but that seems to be only for Named Entity Recognition. Correct, me if I’m wrong.

Essentially, I’m trying to do text generation, and predict the following sequence of characters.

2

Hi, @emmanuelnsanga ,
As you have fine tunned bloom 560M on custom dataset. how is result ,it is better than the bench mark encoder model like (bert , Roberta).

Thanks
Deelip

3

Try LLAMA it’s State-of-the-art…