|
|
--- |
|
|
license: mit |
|
|
datasets: |
|
|
- MuzzammilShah/people-names |
|
|
language: |
|
|
- en |
|
|
model_name: Deep MLP with WaveNet-inspired Architecture |
|
|
library_name: pytorch |
|
|
tags: |
|
|
- makemore |
|
|
- wavenet |
|
|
- convolutional-neural-network |
|
|
- deep-learning |
|
|
- andrej-karpathy |
|
|
--- |
|
|
|
|
|
# Deep MLP with WaveNet-inspired Architecture: Makemore (Part 5) |
|
|
|
|
|
This repository builds upon the 2-layer MLP from previous implementations, making it deeper with a tree-like structure, leading to a convolutional neural network architecture inspired by the WaveNet (2016) from DeepMind. |
|
|
|
|
|
## Overview |
|
|
- This hierarchical architecture is implemented efficiently using causal dilated convolutions (not yet covered). |
|
|
- The project explores `torch.nn` and its functionality, providing insights into deep learning development processes. |
|
|
|
|
|
## Documentation |
|
|
For a better reading experience and detailed notes, visit my **[Road to GPT Documentation Site](https://muzzammilshah.github.io/Road-to-GPT/Makemore-part5/)**. |
|
|
|
|
|
## Acknowledgments |
|
|
Notes and implementations inspired by the **Makemore - Part 5** video by [Andrej Karpathy](https://karpathy.ai/). |
|
|
|
|
|
For more of my projects, visit my [Portfolio Site](https://muhammedshah.com). |
