mbcel

Okay I figured out that the nan's were due the adam optimisation. The default epsilon of 1e-8 is too low and rounded to zero like pointed out [here](https://discuss.pytorch.org/t/adam-half-precision-nans/1765/6). Setting it...

@borisfom : So cutorch.hasFastHalfInstructions() returns false. My GPU is a Titan X Pascal. I tried CUDNN6 with the R6 branch now. It's still not working but now I get a...

Does anyone use CudaHalfTensor successfully with the Titan X Pascal? And if yes, what nvidia driver do you use and which Ubuntu version?

Me too

I agree that this would be a very important feature since especially production ready mlflow servers usually need authentification and especially deep learning experiemnts can usually run for days. In...

No this happens on all threads. No std::cout is working in the callback nor on the main thread that started the camera. Printf is working as expected in both threads....

This ouput is copied from my CLion editor. This is configured to use "utf-8". However the same output is generated in a linux terminal. I am also not sure what...

That's still not clear for me. Independent of the dataset size or used batch size the advantage function should give me positive advantages for things I should do more often...

Any news on this side?

Any news on this? I am experiencing the same issue. I also have a switch inbetween and dhcp mode enabled. I might give it a try with static ip the...