Metadata-Version: 2.1
Name: gmlp-tinygrad
Version: 0.9.2
Summary: An implementation of gated MLPs in tinygrad, as an alternative to transformers.
Author: Ethan Bennett
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE

# Gated MLPs (gMLP) Implementation in Tinygrad

An implementation of gated MLPs in tinygrad, as an alternative to transformers.

## What is a gMLP?

A Gated Multi-Layer Perceptron (gMLP) is a type of deep learning model that uses basic multi-layer perceptrons combined with gating mechanisms to capture spatial interactions between sequence elements, achieving comparable performance to Transformers on various language and vision tasks without relying on self-attention mechanisms.

## Installation

To get started, simply install gmlp_tinygrad using pip:

```bash
pip install gmlp_tinygrad
```

## Usage
```bash
from gmlp_tinygrad import gMLP
```
