https://github.com/buaadreamer/nnplayer

A Toy Nerual Network Framework 自制神经网络框架
https://github.com/buaadreamer/nnplayer

deeplearning deeplearning-framework machine-learning nerual-network nn

Last synced: 9 months ago
JSON representation

A Toy Nerual Network Framework 自制神经网络框架

• Host: GitHub
• URL: https://github.com/buaadreamer/nnplayer
• Owner: BUAADreamer
• Created: 2022-04-16T05:17:19.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2022-04-24T04:18:50.000Z (about 4 years ago)
• Last Synced: 2024-12-21T21:21:57.374Z (over 1 year ago)
• Topics: deeplearning, deeplearning-framework, machine-learning, nerual-network, nn
• Language: Python
• Homepage:
• Size: 23.4 KB
• Stars: 4
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Introduction

This repo is a Nerual Network Framework for learning Deep-Learning and Machine-Learning.

This framework refers to PyTorch's Design and Usage.

## Usage

```python

import nnplayer as npl

import numpy as np

# build net

net = npl.Sequence(npl.Linear(10, 6), npl.Sigmoid(), npl.Linear(6, 4), npl.Sigmoid())

# calculate loss 

loss = npl.MSELoss(net)

# optimizer

optimizer = npl.SGD(net.getParamsList())

# build the dataset

X = np.arange(40).reshape(4, 10)

Y = np.arange(16).reshape(4, 4)

# train

epoch = 100

for i in range(epoch):

    Y_predict = net.forward(X)

    l = loss.loss(Y_predict, Y)

    print(f'epoch{i + 1} loss:{l.mean()}')

    loss.backward()

    optimizer.step()

l = loss.loss(net.forward(X), Y)

print(l.mean())

```