Building a small legal language model from the data up

Estimated read time 1 min read

*Streaming and cleaning two billion tokens, training a 125M transformer, and the things that surprised me along the way.*

 

​ *Streaming and cleaning two billion tokens, training a 125M transformer, and the things that surprised me along the way.*Continue reading on Medium »   Read More AI on Medium 

#AI

You May Also Like

More From Author