Keras - Revision history

Unknown user: /* Flags */

2025-05-11T20:12:21Z

Flags

2018-02-10T18:10:40Z

Github Repos

← Older revision		Revision as of 18:10, 10 February 2018
Line 84:		Line 84:

	* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm		* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm

			===Really Cool Stuff===

			LSTMetallica:
			* https://soundcloud.com/kchoi-research/sets/lstmetallica-drums
			* https://github.com/keunwoochoi/LSTMetallica

	=Notes=		=Notes=

2017-12-02T09:17:42Z

Related Utilities

← Older revision		Revision as of 09:17, 2 December 2017
Line 72:		Line 72:

	===Related Utilities===		===Related Utilities===

			* https://github.com/keplr-io/quiver - convolutional neural net visualization for Keras

	* https://github.com/yusugomori/deeplearning-tensorflow-keras - extensive notes/notebooks on different architectures, divided into chapters/sections		* https://github.com/yusugomori/deeplearning-tensorflow-keras - extensive notes/notebooks on different architectures, divided into chapters/sections
Line 80:		Line 82:

	* https://github.com/sandeep-krishnamurthy/keras-mxnet-benchmarks - Examples for profiling performance of Keras/MXnet		* https://github.com/sandeep-krishnamurthy/keras-mxnet-benchmarks - Examples for profiling performance of Keras/MXnet

	* https://github.com/keplr-io/quiver - convolutional neural net visualization for Keras

	* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm		* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm

2017-10-26T05:02:45Z

Flags

← Older revision		Revision as of 05:02, 26 October 2017
Line 219:		Line 219:
	* (Even earlier) I had initially been using TWO Dense neurons, e.g., Dense(2), to get yes/no. This is not correct!		* (Even earlier) I had initially been using TWO Dense neurons, e.g., Dense(2), to get yes/no. This is not correct!
	* I was trying to use an SGD optimizer... something fancy-pants that I should not have been using		* I was trying to use an SGD optimizer... something fancy-pants that I should not have been using

			=Resources=

			Ugh, a big ugly giant hairy list of links. I'll do my best to keep this maintained.

			==Learning Keras==

			Building a very, very simple neural network with Keras - https://www.pyimagesearch.com/2016/09/26/a-simple-neural-network-with-python-and-keras/

	=Flags=		=Flags=

2017-10-20T04:48:31Z

Errors

← Older revision		Revision as of 04:48, 20 October 2017
Line 84:		Line 84:

	* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm		* https://github.com/joeddav/devol - automating hyperparameter tuning with genetic algorithm

			=Notes=

			I think my in-your-face repository (https://github.com/charlesreid1/in-your-face) does a pretty good job of making notes on how to use Keras as we go.

			Check out the iPython notebooks in that repository.

	=Errors=		=Errors=

2017-10-14T20:49:01Z

Max Pooling 2D Dimension Application

← Older revision		Revision as of 20:49, 14 October 2017
Line 91:		Line 91:
	==Convolutional Neural Networks==		==Convolutional Neural Networks==

	===Max Pooling 2D ~~Dimension Application~~===		===Max Pooling 2D Applied to Wrong Dimensions===

	In a convolutional neural network, the network architecture generally looks like this:		In a convolutional neural network, the network architecture generally looks like this:
Line 141:		Line 141:
	</pre>		</pre>

	This was resulting in the MaxPooling2D layer having an output shape of </code>(4, 8, 32)</code> - the pooling layer was pooling the channels, plus the first dimension of each photo, while leaving the second dimension alone.		This was resulting in the MaxPooling2D layer having an output shape of <code>(4, 8, 32)</code> - the pooling layer was pooling the channels, plus the first dimension of each photo, while leaving the second dimension alone.

	I eventually figured out how to fix this by studying the MaxPooling2D documentation page: https://keras.io/layers/pooling/#maxpooling2d		I eventually figured out how to fix this by studying the MaxPooling2D documentation page: https://keras.io/layers/pooling/#maxpooling2d

2017-10-14T20:17:16Z

Paper/Network Implementations

← Older revision		Revision as of 20:17, 14 October 2017
Line 47:		Line 47:

	* https://github.com/kentsommer/keras-inceptionV4 - Keras implementation of Inception V4 architecture		* https://github.com/kentsommer/keras-inceptionV4 - Keras implementation of Inception V4 architecture

			* https://github.com/flyyufelix/DenseNet-Keras - Keras implementation of DenseNet + ImageNet

	* https://github.com/pengpaiSH/Kaggle_NCFM - Keras to solve Kaggle Nature Conservancy Fisheries Monitoring dataset leaderboard		* https://github.com/pengpaiSH/Kaggle_NCFM - Keras to solve Kaggle Nature Conservancy Fisheries Monitoring dataset leaderboard

	* https://github.com/kylemcdonald/SmileCNN - CNN to detect smiles with Keras		* https://github.com/kylemcdonald/SmileCNN - CNN to detect smiles with Keras

			* https://github.com/titu1994/Super-Resolution-using-Generative-Adversarial-Networks - super resolution with generative adversarial networks

	* https://github.com/snf/keras-fractalnet - FractalNet ultra deep neural networks w/o residuals (M$ paper)		* https://github.com/snf/keras-fractalnet - FractalNet ultra deep neural networks w/o residuals (M$ paper)

2017-10-14T20:11:38Z

Keras

← Older revision		Revision as of 20:11, 14 October 2017
Line 23:		Line 23:

	* https://github.com/fchollet/hualos - Keras total visualization project (Keras RemoteMonitor callbacks - JSON - Flask - C3)		* https://github.com/fchollet/hualos - Keras total visualization project (Keras RemoteMonitor callbacks - JSON - Flask - C3)

			===Keras Forks===

			* https://github.com/MarcBS/keras - Keras fork with additional functionality

			* https://github.com/MarcBS/multimodal_keras_wrapper - wrapper for MarcBS' Keras fork (see prior entry)

			* https://github.com/dmlc/keras - Fork of Keras that supports an MXNet backend

	===Paper/Network Implementations===		===Paper/Network Implementations===

2017-10-14T11:37:55Z

Github Repos

@@ Line 5: / Line 5: @@
 ==Github Repos==
-See https://github.com/charlesreid1/in-your-face - examples of fitting Keras neural networks to the LFW (labeled faces in the wild) dataset.
+===charlesreid1===
-See https://github.com/charlesreid1/lfw_fuel - fork of lfw_fuel repo, which applies kerosene and fuel (Python libraries) for organizing and packaging LFW data in a nice format that makes it easy to load and hand off to Keras.
+* https://github.com/charlesreid1/in-your-face - examples of fitting Keras neural networks to the LFW (labeled faces in the wild) dataset.
-See https://github.com/charlesreid1/circe - specifically the examples using the NIST handwriting digit classification data set. This shows how to utilize Keras to train a neural network to perform dimensionality reduction, and further explores the manifold that the neural network identified for the different digits to better understand the neural network model.
+* https://github.com/charlesreid1/lfw_fuel - fork of lfw_fuel repo, which applies kerosene and fuel (Python libraries) for organizing and packaging LFW data in a nice format that makes it easy to load and hand off to Keras.
 =Errors=

2017-10-14T11:20:45Z

← Older revision		Revision as of 11:20, 14 October 2017
Line 1:		Line 1:
	Neural network package in Python.		Neural network package in Python.

			=Code=

			==Github Repos==

	See https://github.com/charlesreid1/in-your-face - examples of fitting Keras neural networks to the LFW (labeled faces in the wild) dataset.		See https://github.com/charlesreid1/in-your-face - examples of fitting Keras neural networks to the LFW (labeled faces in the wild) dataset.