by aortega on 6/21/23, 8:02 PM with 21 comments
by aortega on 6/21/23, 8:02 PM
by 19h on 7/4/23, 11:15 PM
For instance, obtaining JSON structured data from freetext is a rather impossible task.
That said, for summarization of leq 2k token texts the model performs extremely well.
This is mostly due to its unfiltered nature, where there would be clear biases visible in models like falcon-40b-instruct or other LLaMA derivatives.
by ReactiveJelly on 6/21/23, 8:54 PM
How will you ban bad actors if there's no keys and no registration?
by simonw on 6/21/23, 9:34 PM
by LoganDark on 6/21/23, 10:44 PM
by th0ma5 on 6/21/23, 8:54 PM