https://github.com/mahdizynali/mazenet

Mnist digits multi perceptron neural-network training from scratch with c++ and opencv matrix. (nn with c++)
https://github.com/mahdizynali/mazenet

artificial-intelligence classification cpp cpp-neural-network cpp-training machine-learning mnist mnist-cpp multi-classification neural-network neural-networks opencv preceptron scratch training

Last synced: 6 days ago
JSON representation

Mnist digits multi perceptron neural-network training from scratch with c++ and opencv matrix. (nn with c++)

• Host: GitHub
• URL: https://github.com/mahdizynali/mazenet
• Owner: mahdizynali
• Created: 2023-09-22T09:24:02.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2023-12-03T07:37:19.000Z (12 months ago)
• Last Synced: 2023-12-03T09:27:18.532Z (12 months ago)
• Topics: artificial-intelligence, classification, cpp, cpp-neural-network, cpp-training, machine-learning, mnist, mnist-cpp, multi-classification, neural-network, neural-networks, opencv, preceptron, scratch, training
• Language: C++
• Homepage:
• Size: 7.6 MB
• Stars: 6
• Watchers: 1
• Forks: 0
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# MazeNet

Mnist multi perceptron neural-network training from scratch with c++ and opencv matrix. \

in this repository there are two folder that includes folder contains header files and src contains cpp programs.

## How to train

First clone repository via this command :

```

git clone https://github.com/mahdizynali/MazeNet.git

```

Now try to run shell script easy to use :

```

chmod a+x run.sh

```

```

./run.sh

```

### Run as developer

after that you have to extract dataset from compress file (dataset.tar.xz).

tar.xz is being support by linux & mac systems. \

you would change config.hpp header in order to set your hyperparameter as you need.

```

# define l_rate 0.005

# define total_epochs 10

# define batch_size 64

```

right now inside MazeNet folder follow these commands to build and run training loop :

```

mkdir build && cd build

```

```

make

```

Or try this for a bit faster :

```

make -j`nproc`

```

finally run the program :

```

./maze

```

after training , neuron weights will recive and save into a yml file besides maze program.