bi-att-flow
bi-att-flow copied to clipboard
Tensorboard loss is zero
I'm running the dev branch on tensorflow 1.3.0 and python 3.5.2. The loss at tensorboard is zero throughout the training. Should the tensorboard log work correctly?
I'm running the code with the command
python -m basic.cli --mode train --noload --len_opt --cluster
is there a way to solve this?
Thanks :)