Skip to content

reusable & predefined models built via Keras, which could be easily integrated into your project.

License

Notifications You must be signed in to change notification settings

Marcnuth/Keras-Models

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Keras Models Hub

PyPI - Downloads PyPI

This repo aims at providing both reusable Keras Models and pre-trained models, which could easily integrated into your projects.

Install

pip install keras-models

If you will using the NLP models, you need run one more command:

python -m spacy download xx_ent_wiki_sm

Usage Guide

Import

import kearas_models

Examples

LinearModel

DNN

CNN

from keras_models.models import CNN

# fake data
X = np.random.normal(0, 1.0, size=500 * 100 * 100 * 3).reshape(500, 100, 100, 3)
w1 = np.random.normal(0, 1.0, size=100)
w2 = np.random.normal(0, 1.0, size=3)
Y = np.dot(np.dot(np.dot(X, w2), w1), w1) + np.random.randint(1)

# initialize & train model
model = CNN(input_shape=X.shape[1:], filters=[32, 64], kernel_size=(2, 2), pool_size=(3, 3), padding='same', r_dropout=0.25, num_classes=1)
model.compile(optimizer='adam', loss=mean_squared_error, metrics=['mae', 'mse'])
model.summary()

model.fit(X, Y, batch_size=16, epochs=100, validation_split=0.1)

SkipGram

WideDeep

VGG16_Places365 [pre-trained]

This model is forked from GKalliatakis/Keras-VGG16-places365 and CSAILVision/places365

from keras_models.models.pretrained import vgg16_places365
labels = vgg16_places365.predict(['your_image_file_pathname.jpg', 'another.jpg'], n_top=3)

# Example Result: labels = [['cafeteria', 'food_court', 'restaurant_patio'], ['beach', 'sand']]

Models

  • LinearModel
  • DNN
  • WideDeep
  • TextCNN
  • TextDNN
  • SkipGram
  • ResNet
  • VGG16_Places365 [pre-trained]
  • working on more models

Citation

WideDeep

Cheng H T, Koc L, Harmsen J, et al. 
Wide & deep learning for recommender systems[C]
Proceedings of the 1st workshop on deep learning for recommender systems. ACM, 2016: 7-10.

TextCNN

Kim Y. 
Convolutional neural networks for sentence classification[J]. 
arXiv preprint arXiv:1408.5882, 2014.

SkipGram

Mikolov T, Chen K, Corrado G, et al. 
Efficient estimation of word representations in vector space[J]. 
arXiv preprint arXiv:1301.3781, 2013.

VGG16_Places365

Zhou, B., Lapedriza, A., Khosla, A., Oliva, A., & Torralba, A.
Places: A 10 million Image Database for Scene Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence

ResNet

He K, Zhang X, Ren S, et al. 
Deep residual learning for image recognition[C]
Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.

Contribution

Please submit PR if you want to contribute, or submit issues for new model requirements.