You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2018/09/11 23:00:29 UTC

[GitHub] stu1130 commented on a change in pull request #12419: [MXNET-580] [WIP] Add SN-GAN example

stu1130 commented on a change in pull request #12419: [MXNET-580] [WIP] Add SN-GAN example
URL: https://github.com/apache/incubator-mxnet/pull/12419#discussion_r216850573
 
 

 ##########
 File path: example/gluon/sn-gan/README.md
 ##########
 @@ -0,0 +1,40 @@
+# Spectral Normalization GAN
+
+This example implements [Spectral Normalization for Generative Adversarial Networks](https://arxiv.org/abs/1802.05957) based on [CIFAR10](https://www.cs.toronto.edu/~kriz/cifar.html) dataset.
+
+## Usage
+
+Example runs and the results:
+
+```python
+python train.py --use-gpu --data-path=data
+```
+
+* Note that the program would download the CIFAR10 for you
+
+`python train.py --help` gives the following arguments:
+
+```bash
+optional arguments:
+  -h, --help            show this help message and exit
+  --data-path DATA_PATH
+                        path of data.
+  --batch-size BATCH_SIZE
+                        training batch size. default is 64.
+  --epochs EPOCHS       number of training epochs. default is 100.
+  --lr LR               learning rate. default is 0.0001.
+  --lr-beta LR_BETA     learning rate for the beta in margin based loss.
+                        default is 0.5s.
+  --use-gpu             use gpu for training.
+  --clip_gr CLIP_GR     Clip the gradient by projecting onto the box. default
+                        is 10.0.
+  --z-dim Z_DIM         dimension of the latent z vector. default is 100.
+```
+
+## Learned Spectral Normalization
+
+![alt text](https://github.com/taki0112/Spectral_Normalization-Tensorflow/blob/master/assests/sn.png)
+
 
 Review comment:
   I am running the model with Xavier initializer and will update the image if it's better

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services