3 ms·

What would be the most performant way to run a inference using RWKV? Do you have and speed comparison to a similar sized transformer? I have a task(OCR cleanin

by bratao 2y ago

What would be the most performant way to run a inference using RWKV? Do you have and speed comparison to a similar sized transformer? I have a task(OCR cleaning) that I´m evaluating faster options and look like RWKV would be a nice alternative.