Leon Unger

@matthew-mcateer Actually there was no message at all apart from the error when using `with tf.device('/gpu:0'):` Let me know if there is a way how I can produce a stack...

@matthew-mcateer FYI There is a significant difference in terms of speed when running TFP in session or eager mode Example is the "A/B testing" example in Chp 2 on my...

Some more descriptive statistics on the usage of ressources: 1. with tf.device('cpu:0'): 80-90% CPU usage and some GPU around 30% 1. without decleration of devices: 50% CPU and the same...

Hi @matthew-mcateer, do you already have an idea for "CPU-GPU-issue"? Cheers

Hi @matthew-mcateer, is there news? Thanks in advance

@matthew-mcateer I just updated to tf 1.13.1 with Cuda 10 and cuDNN 7.5 and tf-probability 0.6. Unfortunately it's still true that the "gpu-mode-sampling" is much slower (increasing in sample size)...

@matthew-mcateer I give an update on the issue: After switching to TF 2.0_nightly and TFP_nightly there is a changed behavior: - eager mode is default, hence i leave it this...

@Makan-Ar I think you are right. Each coin-flip is drawn from a Bernoulli distribution. However, the consecutive draws by which we would like to check the fairness of the coin...

@jgm Thank you for your fast response. Please find attached 4 docx files which show the problem: [rmd2docx_pandoc_with_templates_2_10_1_vs_2_19_2.zip](https://github.com/jgm/pandoc/files/9438378/rmd2docx_pandoc_with_templates_2_10_1_vs_2_19_2.zip) - rmd2docx_pandoc_with_default_template.docx → rmd2docx with default template - MyListTemplate.docx → the designed...

@jgm I see. Hence, the correct label for this issue would be **feature request**. As a matter of fact, being able to change the default list styling is a nice...