The codes of paper "Long Text Generation via Adversarial Training with Leaked Information" on AAAI 2018. Text generation using GAN and Hierarchical Reinforcement Learning.
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 done to it. I have a text dataset, but replacing realtrain_Cotra doesn't work. I thought it needed to be onehot encoded, but that didn't work. I thought maybe it just needed to be label encoded but that doesn't work. I converted the label encoded integer array into a simple string to try and closely match the kind of data you have in your realtrain_cotra but then it gives me a divide by zero error.
epoch: 0
Traceback (most recent call last):
File "Main.py", line 285, in
main()
File "Main.py", line 177, in main
gen_data_loader.create_batches(positive_file)
File "/content/LeakGAN/Image COCO/dataloader.py", line 22, in create_batches
self.sequence_batch = np.split(np.array(self.token_stream), self.num_batch, 0)
File "/usr/local/lib/python2.7/dist-packages/numpy/lib/shape_base.py", line 847, in split
if N % sections:
ZeroDivisionError: integer division or modulo by zero
what in the world do you do to get this system working on custom data?
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 done to it. I have a text dataset, but replacing realtrain_Cotra doesn't work. I thought it needed to be onehot encoded, but that didn't work. I thought maybe it just needed to be label encoded but that doesn't work. I converted the label encoded integer array into a simple string to try and closely match the kind of data you have in your realtrain_cotra but then it gives me a divide by zero error.
what in the world do you do to get this system working on custom data?