LeakGAN icon indicating copy to clipboard operation
LeakGAN copied to clipboard

The codes of paper "Long Text Generation via Adversarial Training with Leaked Information" on AAAI 2018. Text generation using GAN and Hierarchical Reinforcement Learning.

Results 21 LeakGAN issues
Sort by recently updated
recently updated
newest added

If we want to run the model on a custom dataset, how should we do this?

`def rescale( reward, rollout_num=1.0): reward = np.array(reward) x, y = reward.shape ret = np.zeros((x, y)) for i in range(x): l = reward[i] rescalar = {} for s in l: rescalar[s]...

/home/lishaomei/anaconda2/bin/python "/home/lishaomei/LeakGAN-master/Synthetic Data/Main.py" 20 WARNING:tensorflow:From /home/lishaomei/LeakGAN-master/Synthetic Data/Discriminator.py:92: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version. Instructions for updating: Future major versions of TensorFlow will allow...

Traceback (most recent call last): File "/Users/zhouye/Documents/dtwave/LeakGAN/run.py", line 14, in main() File "/Users/zhouye/Documents/dtwave/LeakGAN/Image_COCO/Main.py", line 161, in main batch_size=BATCH_SIZE,hidden_dim=HIDDEN_DIM,start_token=START_TOKEN,goal_out_size=GOAL_OUT_SIZE,goal_size=GOAL_SIZE,step_size=4,D_model=discriminator) File "/Users/zhouye/Documents/dtwave/LeakGAN/Image_COCO/LeakGANModel.py", line 284, in __init__ self.pretrain_manager_updates = pretrain_manager_opt.apply_gradients(pretrain_manager_grads_vars) File "/Users/zhouye/miniconda/envs/py36/lib/python3.6/site-packages/tensorflow/python/training/optimizer.py", line...

I ran the codes which In the path(\LeakGAN-master\Image COCO\Main.py)(LeakGAN-master\Image COCO\eval_bleu.py). but I can't get the same results as the paper. I just changed the python's version(3.6) and tensorflow-gpu 's version(1.8.0)....

I've been trying to get your LeakGAN to work for some time now but I just don't understand what format your realtrain_cotra file is or what kind of preencoding you've...

I'm new to LeakGAN or SeqGAN or TextGAN. I know GAN is to generate text and let discriminator un-judge-able to real text and gen-text. LM(language model) is the task of...

![image](https://user-images.githubusercontent.com/4702353/54328665-c13bf600-4649-11e9-92a4-de5a8eaea165.png) And `leakgan.x` is part of worker loss: ![image](https://user-images.githubusercontent.com/4702353/54329100-4f64ac00-464b-11e9-92a3-bbf0bbeb0fd9.png) Thank you!!! Thank you!!! Thank you!!! @wnzhang @CR-Gjx