Home▸AI & Machine Learning▸FedAvg Loss-Curve Drift: 2D Federated Model Aggregation

🌐 FedAvg Loss-Curve Drift: 2D Federated Model Aggregation

Interactive 2D federated-learning simulator: watch five clients run local gradient steps on a shared 1D loss curve and drift toward their own non-IID optimum, then FedAvg's sample-weighted average pulls the global model back, round after round.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version

⚙ Under the hood

Interactive 2D federated-learning simulator: five clients report scalar weight updates onto a single shared loss-curve graph, drifting toward their own non-IID optimum during local training before FedAvg's sample-weighted average pulls the global model back — with a live convergence chart and aggregation-weight bar alongside the loss curve.

federated learningFedAvgmodel aggregationnon-IID dataclient driftmachine learningconvergence

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install