On what amount of English tokens was the model trained on? Is it above 2.6 Trillion tokens?
· Sign up or log in to comment