Basically removing the self-labeling part? That would make it clearer.
This would simply the spec and makes indexing easier for relays, but we loose the data atomicity
Please Login to reply.
Is it a dataset to train LLMS?
Actually it could be that
dan dropping bombs ;)