Our @ed asked a full room how tightly coupled should the training dataset and the trained #AI #ML model be for an "Open" ML? Do we absolutely need the original dataset or would it be enough to have a very good description of it in order to modify a model?
The #FOSSY group discussion was amazing.
Conversation
Notices
-
Embed this notice
Open Source Initiative :osi: (osi@social.opensource.org)'s status on Saturday, 15-Jul-2023 09:03:16 JST Open Source Initiative :osi: - Haelwenn /элвэн/ :triskell: likes this.