bitgeist 0.1.3

bitgeist is a Python library to convert deep neural network models into c source code.

bitgeist

bitgeist is a Python library to convert deep neural network models into c source code.

Installation

Use the package manager pip to install bitgeist.

pip install bitgeist

Usage

import bitgeist as bg

# transform a pytorch model to c code
print(bg.transform(model))

Example:

    ...
    class MnistModel(nn.Module):

        def __init__(self, downsample=None):
            super(MnistModel, self).__init__()

            inputs = 28 * 28
            hidden = 120
            output = 10

            self.l1 = DenseBlock(inputs, hidden, downsample=downsample)
            self.action = nn.PReLU(hidden)
            self.l2 = DenseBlock(hidden, output, downsample=downsample)

        def forward(self, x):
            out = self.l1(x)
            out = self.action(out)
            out = self.l2(out)
            return out

    model = MnistModel(nn.Dropout(0.2)).to(device)

    # train the model 
    
    ...
    
    print("total training time = {} hours".format(training_time))

    print(bg.transform(model)) #print or write to file ...
   ...

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT