Thanks for implementing the work.
You mentioned in the readme that there are a number of issues with the RIN architecture. Could you provide some more details?
What do u think might be the reason that self-conditioning and the architecture design enable strong FIDs without using guidance?
Thanks for implementing the work. You mentioned in the readme that there are a number of issues with the RIN architecture. Could you provide some more details? What do u think might be the reason that self-conditioning and the architecture design enable strong FIDs without using guidance?