Spaces:

---

bigscience-data

/

README

like 0

Running

Hi there,

I would like to use Bloom in my research. I was wondering whether there is a way to reconstruct the tokenised pre-training data in the order seen by the model.

Thanks,
Pietro