We just launched Deep-ML Labs, a new section on Deep-ML

Each Lab is a hands-on challenge where you implement parts of real models, activations, layers, optimizers, normalization, even small transformers — directly in PyTorch (and sometimes NumPy).

You can test your implementations on actual datasets, see how your results compare to others, and understand what’s really happening under the hood of deep learning frameworks.

let me know what you guys think and if we should work on making more of these questions