I don't write Python, but it looks like I can read it. If you could create a github repo with an interface for the missing pieces, I'm confident we can adapt the Python parser/validator (we already have) to expose set.mm pretty much however you need it.
My main concern is that the "uncompressed" version of set.mm (that I expect to be the more effective, for a learning machine) would be enormous, thus it has to be "created" on the fly (not a big deal, anyway). Il giorno lunedì 4 marzo 2024 alle 07:56:49 UTC+1 [email protected] ha scritto: *Would you or a group of folks be interested in collaborating on a follow-up paper to use this framework to build a mm GPT?* -- You received this message because you are subscribed to the Google Groups "Metamath" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/metamath/eb571275-4d3e-4f11-a333-7d727db74438n%40googlegroups.com.
