I am building a tool to train machine learning enhanced filters against kind 1 (or eventually any content containing) events. Currently I spend time (hours) training and end up with a payload of blocked npubs and a bayes ML model as json. My open question is how do I and my users share this payload with the community over nostr in a sustainable fashion? I wish to collaborate more and I'm looking for a nuanced answer than "you should charge people sats on a marketplace!" I already spent time on a tool to share git patches. Seems like I could sell these patches or the whole json payload in the vending machine. Do you have code somewhere or anything concrete to get me started? cafe-society.news
Discussion
I’ll publish my repo tonight so you can see how to sell it
I’ll experiment with buying you the data from the nostr:npub1w0rthyjyp2f5gful0gm2500pwyxfrx93a85289xdz0sd6hyef33sh2cu4x npub
What git patches are you referring to? Can I see an example?
I have no examples on hand and the tool is POC terminal only. I'll respond with example in a few hours, after the breakfast club. meanwhile this is what I plan to implement in javascript at some point. ( rip off of git-ssb )
https://github.com/colealbon/git-nostr. converts the output of git format patch to base64
I want to sell the JSON output of https://winkjs.org/wink-naive-bayes-text-classifier/
[{"considerOnlyPresence":true,"smoothingFactor":0.5},{"promote":1,"suppress":1},{"promote":{"celo":1,"blockchain":1,"propo":1,"return":1,"ethereum":1,"ecosystem":1,"transit":1,"l2":1,"accord":1,"clab":1,"includ":1,"leverag":1,"op":1,"stack":1,"architectur":1},"suppress":{"buy":1,"dip":1,"record":1,"bitcoin":1,"suppli":1,"k":1,"chain":1,"data":1,"confirm":1,"signif":1,"point":1,"interest":1,"btc":1,"accuml":1}},{"promote":15,"suppress":14},["celo","blockchain","propo","return","ethereum","ecosystem","transit","l2","accord","clab","includ","leverag","op","stack","architectur","buy","dip","record","bitcoin","suppli","k","chain","data","confirm","signif","point","interest","btc","accuml"]]