wassname/keras-js
1
0 Fork 0
mirror of https://github.com/wassname/keras-js.git synced 2026-10-03 12:31:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

update readme

Browse Source
This commit is contained in:
Leon Chen committed 2016-10-24 11:04:17 -04:00
1 parent b459b78392
commit 8e7de948c1
1 file changed
+15 -7
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+15 -7
View File
@@ -1,9 +1,20 @@
# [Keras.js](https://transcranial.github.io/keras-js)
Run [Keras](https://github.com/fchollet/keras) models (trained using Tensorflow backend) in your browser, with GPU support. Models are serialized directly from the Keras JSON-format configuration file and associated HDF5 weights. GPU support is powered by WebGL through [weblas](https://github.com/waylonflinn/weblas).
Run [Keras](https://github.com/fchollet/keras) models (trained using Tensorflow backend) in your browser, with GPU support. Models are created directly from the Keras JSON-format configuration file, using weights serialized directly from the corresponding HDF5 file.
Inspiration is drawn from a number of deep learning / neural network libraries for JavaScript and the browser, including [ConvNetJS](https://github.com/karpathy/convnetjs), [synaptic](https://github.com/cazala/synaptic), [brain](https://github.com/harthur/brain), [CaffeJS](https://github.com/chaosmail/caffejs), [MXNetJS](https://github.com/dmlc/mxnet.js).
Tensor operations are extended on top of the [ndarray](https://github.com/scijs/ndarray) library. GPU support is powered by WebGL through [weblas](https://github.com/waylonflinn/weblas). The focus here is on inference only.
### [Interactive Demos](https://transcranial.github.io/keras-js)
<p align="center">
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/mnist-cnn.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/resnet50.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/inception-v3.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/imdb-bidirectional-lstm.png" height="120" width="auto" /></a>
</p>
- Basic Convnet for MNIST
- Convolutional Variational Autoencoder, trained on MNIST
@@ -12,14 +23,11 @@ Run [Keras](https://github.com/fchollet/keras) models (trained using Tensorflow
- Inception V3, trained on ImageNet
- Xception V1, trained on ImageNet
- Bidirectional LSTM for IMDB sentiment classification
<p align="center">
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/mnist-cnn.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/resnet50.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/inception-v3.png" height="120" width="auto" /></a>
<a href="https://transcranial.github.io/keras-js"><img src="demos/assets/imdb-bidirectional-lstm.png" height="120" width="auto" /></a>
</p>
*planned*: Char-RNN, SqueezeNet
### Usage