Caffe CIFAR-10 Build and Training

This skill provides procedural guidance for building the Caffe deep learning framework from source and training models on the CIFAR-10 dataset.

When to Use This Skill

• Building Caffe from source on Ubuntu/Debian systems
• Training CIFAR-10 or similar image classification models with Caffe
• Configuring Caffe for CPU-only execution
• Troubleshooting Caffe build and dependency issues

Critical Requirements Checklist

Before starting, identify ALL requirements from the task specification:

1. Execution mode: CPU-only vs GPU (affects solver configuration)
2. Iteration count: Specific number of training iterations required
3. Output files: Where training logs and models should be saved
4. Model checkpoints: Which iteration's model file is expected