{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/5d04822a0933427d971d320f64687730\" frameborder=\"0\" width=\"1890\" height=\"1417\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1417,"width":1890,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1417,"thumbnail_width":1890,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/5d04822a0933427d971d320f64687730-9804012f5b000e1f.gif","duration":242.372,"title":"AI Model Training with WeightsLab [dev]","description":"In this video, I introduced WeightsLab, a visual control panel designed for AI developers to inspect, modify, and enhance deep neural networks during training. Using a simple example of a neural network trained on the MNIST dataset, I demonstrated how to manipulate data by focusing on high-loss examples, which helped us reduce the dataset by around 95%. I also showed how to freeze layers and add capacity to improve the model, resulting in an accuracy increase from 86% to 92% with just 5% of the initial dataset. I encourage you to explore WeightsLab to see how it can transform your approach to AI model training and debugging."}