Yasuhiro Fujita
Yasuhiro Fujita
@PaleNeutron Thanks! My workaround for now is to make an executable shell script like `google-chrome --no-sandbox "$@"` and specify the `chrome_path` argument of `dataframe_image.export` to it.
I am facing the same issue. `table_conversion="chrome"` works with `set_caption` but matplotlib does not.
Thanks for raising the issue. pfrl==0.4.0 has been released, which should solve it.
👍 for this feature. TensorFlow's `MixtureSameFaimily` has a `reparameterize` option using using implicit reparameterization gradients. https://www.tensorflow.org/probability/api_docs/python/tfp/distributions/MixtureSameFamily
The changes will probably not break anything, but can you elaborate on why you suggest them? If it is just for consistency, I think it is better to force recorded...
> To be clear, we do indeed force them to be floats, right? We simply do loss.detach().cpu().numpy() before casting them to floats. Ah sorry I missed `float` around it. Right,...
> Perhaps I can replace instances of detach().cpu().numpy() with tensor.item()? And check the tests and maybe some runs and so forth? Sounds good. I think it is ok to write...
@qgallouedec Can you review this again?