Build and review a complete Deep Learning project.
Project brief
Required concepts
- Tensor contracts: shape, dtype, device and mask
- Logits, cross-entropy and gradients: align the training calculation
- Batching and class sampling: know the population the optimizer sees
- Image augmentation: split originals first and preserve the label
- Training and validation modes: measure the model you will serve
- Checkpoint recovery: save optimizer state and the run boundary
- Inference contracts: preserve preprocessing and measure tail latency
Completion standard
Submit inputs, runnable work, measured output, failure-case results and a short decision note.