Update README.md

README.md

@@ -30,6 +30,12 @@ One command
 $ lightweight_gan --data ./path/to/images --image-size 512
 ```
 
+--data : Path to the folder where train images are present 
+--image-size : Size of the image to train
+--batch-size : Reduce batch size if you are getting GPU error
+
+While training the model, you can see generated images in results folder
+
 Model will be saved to `./models/{name}` every 1000 iterations, and samples from the model saved to `./results/{name}`. `name` will be `default`, by default.
 
 ## Training settings
@@ -45,6 +51,33 @@ $ lightweight_gan \
     --num-train-steps 200000
 ```
 
+
+## Generating
+
+Once you have finished training, you can generate samples with one command. You can select which checkpoint number to load from. If `--load-from` is not specified, will default to the latest.
+
+```bash
+$ lightweight_gan \
+  --name {name of run} \
+  --load-from {checkpoint num} \
+  --generate \
+  --generate-types {types of result, default: [default,ema]} \
+  --num-image-tiles {count of image result}
+```
+
+Example : lightweight_gan --name default --load-from 5 --generate --generate-types default --num-image-tiles 100
+
+
+After run this command you will get folder near results image folder with postfix "-generated-{checkpoint num}".
+
+You can also generate interpolations
+
+```bash
+$ lightweight_gan --name {name of run} --generate-interpolation
+```
+
+
+
 ## Augmentation
 
 Augmentation is essential for Lightweight GAN to work effectively in a low data setting