green flag-change training set (does not reset AI) c-reset AI (clears learning progress) hold space-train AI turbowarp ver- https://turbowarp.org/841639467?fps=60&hqpen&stuck&limitless
Testing a new instance-less training method(no batches) that tests adjustments to each weight and bias and uses a new variable error calculator (no longer # of correct vs wrong). This AI should be better than v2 at learning datasets.
