User Portlet
| Discussions |
|---|
| It looks like there was a bug in the HistogramDistribution code that was introduced in version 10.4 and fixed in version 11.2. I'm afraid I don't know of a workaround for the affected versions. And the code I used was: data =... |
| Do you have the very latest Nvidia drivers installed? In System Preferences > CUDA does it say that a newer driver is available? |
| CUDALink is entirely separate from the NeuralNetworks GPU computing via CUDA. Neither depends on the other in any way. CUDALink still needs to be updated for CUDA Toolkit 8. |
| I asked one of our doc developers, and this was their response (slightly edited): > Looks like the docs tools in WB don't support it. This code will make a small palette that does mostly the same thing as > what we have in our internal tools. ... |
| Sorry, my comment was a bit vague. What I meant is that the doc building code in Workbench (10.1.822) is based on the code used to build the official M11.1 documentation. So when that code was modified to omit the tutorial links, that trickled down... |
| The original issue here is a FE warning, not a crash. If you're experiencing an application crash, please contact Wolfram support with all details on what triggered the crash on your machine. |
| Thank you for posting this idea. We are discussing this internally. |
| Ah, the line with "nvcc fatal : >" is key. The issue is that CUDA Toolkit 7.5 doesn't support the very latest clang C compiler (which Xcode installs). So, unless I'm mistaken, this is the situation: (a) Before you installed Xcode, the problem was... |
| I need to check with the developers, but what I think is happening is that the internal checks that guard against exhausting GPU memory during training are failing. Normally it should immediately see that training wouldn't work and bail out (giving... |
| I don't think this can be updated via a paclet. |