Hi, I'm working on a project about achieving consistency across code repositories, and I'm really interested in it, especially with AI getting stronger these days. What I'm particularly curious about is how to create the right unified abstraction across different code repositories. I saw that the pith-train framework advertises itself as AI-friendly, and I'm really curious how to make a proper abstraction here so that I can automate the transfer of training the same mathematical logic model across different frameworks, and keep it as formal as possible, rather than relying entirely on AI.
Hi, I'm working on a project about achieving consistency across code repositories, and I'm really interested in it, especially with AI getting stronger these days. What I'm particularly curious about is how to create the right unified abstraction across different code repositories. I saw that the pith-train framework advertises itself as AI-friendly, and I'm really curious how to make a proper abstraction here so that I can automate the transfer of training the same mathematical logic model across different frameworks, and keep it as formal as possible, rather than relying entirely on AI.