Affine Tackles Model Spam With Fee Proposal

Affine's subnet faces queue congestion from 70+ merged/unmodified models submitted without real training. Team proposes introducing submission fees to filter spam and reduce computational waste. DNS API issues reported. Top miners reportedly left due to network quality concerns, particularly overfitting in the liveweb environment.

• •70+ models queued since new sampling system; mostly merged or weight-changed, not trained
• •Proposed submission fee to prevent spam while keeping queue clean
• •DNS server timeout errors affecting Affine API access
• •Top performers left subnet citing overfitting issues in broken environments