Skip to main content

Chainer from/to NNOIR Converter

Project description

nnoir-chainer

Chainer Model from/to NNOIR converter

ATTENTION: This library goes into the maintenance phase too.

Install

pip install nnoir-chainer

Example

Import NNOIR

import chainer
from nnoir_chainer import NNOIRFunction
m = NNOIRFunction('nnoir_file_path')
x = chainer.Variable(np_array)
with chainer.using_config('train', False):
    y = m(x)
    print(y)

Export NNOIR

m = model.CNN()
chainer.serializers.load_npz('cnn.model', L.Classifier(m))
with chainer.using_config('train', False):
    x = chainer.Variable(np.zeros((1, 28*28)).astype(np.float32))
    y = m(x)
    g = nnoir_chainer.Graph(m, (x,), (y,))
    result = g.to_nnoir()
    with open('model.nnoir', 'w') as f:
        f.buffer.write(result)

These layers are supported by nnoir-chainer exporter.

  • chainer.links
    • BatchNormalization
    • Bias
    • Linear
    • Convolution2D (DepthwiseConvolution2D, DilatedConvolution2D)
    • Scale
    • Swish
  • chainer.function
    • Add
    • AddConstant
    • AveragePooling2D
    • ClippedReLU
    • Concat
    • Dropout
    • ELU
    • LeakyReLU
    • MaxPooling2D
    • Mul
    • MulConstant
    • Pad
    • ReLU
    • Reshape
    • Sigmoid
    • Softmax
    • Sub
    • Sum
    • Tanh
    • Transpose
    • Unpooling2D

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

nnoir-chainer-1.1.3.tar.gz (17.2 kB view hashes)

Uploaded Source

Built Distributions

nnoir_chainer-1.1.3-py3.7.egg (145.2 kB view hashes)

Uploaded Source

nnoir_chainer-1.1.3-py3-none-any.whl (57.0 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page