Summary: 13 instances, 13 unique

Text	Count
# Gen optimizer TODO: URGENT	1
cfg.DISCRIMINATOR.beta_max = 100.  # TODO: temperature decay	1
# TODO: low priority could potentially make forward_generate a static func?	1
self.cfg.DISCRIMINATOR.batch_chunk * self.cfg.DISCRIMINATOR.sample_chunks_mem)).backward()  # TODO CNN WGAN-GP	1
# TODO: investigate training tricks for dis different clip?	1
# TODO: When training gen do not pass only context into dis (since no grads anyway)	1
# TODO: reload epoch and scheduler	1
# TODO: can we move handling of mems to inside transformer_gan?	1
# TODO: dis training messes up memory structure maintained during batch loading	1
# TODO Add more options	1
# TODO: do not loop over context	1
# TODO: Can we retain sub graph after calling backward?	1
# TODO Think about how to enable masking when we reach BOS.	1