Tzu-Wei Huang

Interesting, so are those `same_embedding` lies in 2D space?

Can you post the whole screenshot of tensorboard page? I can rotate the view without a problem. Or just sharing `same_embedding` is also helpful.

Hi, I changed `writer.add_image(name, grid.numpy(), 1)` to `writer.add_image('img_'+name, grid.numpy(), 1)` and everything works. It appears that it's a name conflict problem and will need further investigation. (the `name` was used...

I tried make the same conflict in `demo.py` and the situation is the same. There is no problem in your code.

Looks like you didn't install tensorflow correctly. running `tensorboard --logdir runs` should get you some server information in the console.

Yes, tensorboardX is under heavy reconstruction in hope that it supports different backends in the future.

Yes, you can specify the `logdir` when constructing `SummaryWriter`, where `logdir` is the folder name of previous run. check examples/demo_purge.py

Hi, If you are using TensorFlow with eager mode, you should able to get the value in numpy array. Once you have the numpy value, you can use tensorboardX to...

Hi, which version did you use? from pip or from source? On Fri, Jan 18, 2019 at 3:34 PM Tim Zaman wrote: > tensorboardX==1.6.0 > > See title, eg tensorboard...

@jandevries123 Will it still freeze if you change `predicted_action[0][x]` to a scalar and `self.gstep` to a increasing scalar?