Name		Name	Last commit message	Last commit date
parent directory ..
gen_img		gen_img
model		model
mrgan_model.py		mrgan_model.py
mrgan_tb.png		mrgan_tb.png
mrgan_train.py		mrgan_train.py
readme.md		readme.md

readme.md

Mode Regularized Generative Adversarial Networks

Loss Function

used GAN + (Enc) loss at D/G nets.
used Adv + MSE loss at Encoder.

Architecture Networks

Same as MRGAN paper.

DIFFS	MRGAN Paper	ME
Weight initializer	`normal dist`	`HE initializer`
z dim	`100`	`128`

HE Initializer parameters : (factor = 1, FAN_AVG, uniform)

Tensorboard

Elapsed Time : 5h 24m 2s with GTX 1060 6GB x 1

Result

Name	Global Step 8k	Global Step 16k	Global Step 32k
MRGAN

To-Do