Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Dive into the final part of a comprehensive hands-on coding workshop focused on advanced deep learning techniques for satellite imagery. Learn remote debugging with Weights & Biases, save and reload models with custom metrics and loss functions, generate activation and gradient outputs with heatmaps, and deploy a model serving application on Hugging Face. Explore topics such as remote debugging, model persistence, custom image prediction, activation visualization, and creating a prediction app with Gradio UI. Follow along with step-by-step instructions and access the complete source code on GitHub to enhance your skills in processing and analyzing satellite image data using deep learning methods.
Syllabus
- Content Starts
- What is covered?
- Tutorial Starts
- Tutorial Starts
- Remote Debugging
- Model Save/Retrieve
- Prediction with custom image
- Activation Output
- Activation Heatmap
- Prediction app
- App Code at Colab
- Gradio UI
- Hugging Face Deployment
- Source Code at GitHub
Taught by
Prodramp