Is this instruction following model?
#1
by
rjmehta
- opened
The Avg looks pretty impressive. Is this pretrained, finetuned, or instruction tuned model?
Yeah it's finetuned on a large set of diverse tasks (~400 million tokens x 3 epochs). Thanks for your interest in our model! Also check out DiscoResearch/DiscoLM-120b for even stronger benchmark performance :)
bjoernp
changed discussion status to
closed