Metadata-Version: 2.4
Name: supernet_duo
Version: 0.1.7
Summary: Supernet_Duo: A extremley quick and simple neural network classifier, and prediction model now with option to return class or percent sure of that class when using the classifier.
Author-email: Bodhi <bodhishanbhag@gmail.com>
License: MIT License
        
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file

# Supernet_Duo

An extremely powerful, lightweight, and simple text classification model, consisting of two neurons each with assigned jobs and a parabolic non-linear activation function. The weight of the connection between the two neurons is variable, and the model is trained and run instantly on the data provided in a dictionary.

## Installation
```bash
pip install supernet_duo
```

## Usage
```python
from supernet_duo import network, predict_net
#Data that the model is trained on
mydata = {'123':'Class1','321':'Class2','01234':'Class1'}
#Text that is run through the model
input_string = '3210'
#Stregnth of the connection between neurons
weight = 4
#Run function
print(network(input_string,mydata,weight,'Class'))
#Output: Class2
#Run function to return the confidence as a float between 0.5 and 1
print(network(input_string,mydata,weight,'Float'))
#Output: 0.9630753667172484 
training_data = ["hello world","hello there"]
input_string = "h"
output_size = 100
print(predict_net(input_string,output_size,training_data, layer1=1000, layer2=1000, window_size=2))
#output~: hleolworldwtl dhhtolddt therehwltwd eht hw etrere dtethello therello therehrweeedeehoeodere worldeeer


```
