All Katas - ML Katas

Custom `nn.Module` with a Non-standard Initialization

this month medium (<10 mins) | initialization custom module weights bias

Create a **custom `nn.Module`** for a simple feed-forward layer. Instead of the default PyTorch initialization, you'll apply a specific, non-standard initialization scheme. For example, you could...

Implementing a Custom `nn.Module` for a Gated Recurrent Unit (GRU)

this month medium (<1 hr) | rnn gru custom module recurrent

Implement a **custom GRU cell** as a subclass of `torch.nn.Module`. Your implementation should handle the reset gate, update gate, and the new hidden state computation from scratch, using...

Implementing Layer Normalization from Scratch

this month medium (<30 mins) | training custom module normalization layer

Implement **Layer Normalization** as a custom `torch.nn.Module`. Unlike `BatchNorm`, `LayerNorm` normalizes across the features of a single sample, not a batch. Your implementation should...