| 29 Nov 2022 |
 tpw_rules | hexa: running a nixpkgs-review cycle including CUDA stuff. planning to try the test suite too. expect something by tomorrow evening but i don't expect any major issues | 05:05:55 |
| tpw_rules | thank you for your patience | 05:06:53 |
 Samuel Ainsworth | draft CUDA 11.8 upgrade: https://github.com/NixOS/nixpkgs/pull/203658 | 20:49:35 |
| Samuel Ainsworth | for some reason jaxlib and tensorflow are not building with 11.8... anyone have any ideas? | 20:49:56 |
| tpw_rules | i thought Someone S had a draft too and he got similar errors | 20:58:37 |
| Samuel Ainsworth | oh that's enitrely possible... i was not aware | 23:50:55 |
| Samuel Ainsworth | yup, you're totally right: https://github.com/NixOS/nixpkgs/pull/200435 | 23:51:40 |
| Samuel Ainsworth | i'm really confused why we're seeing these errors... they seem to indicate that the directory structure changed between 11.7 -> 11.8 | 23:52:23 |
| 1 Dec 2022 |
 @box1:matrix.org | I'm trying to package dgl-cu116(dgl with cuda support) and it fails to find rpath for libtorch_cuda_cpp.so and libtorch_cuda_cu.so.
After some searching, those files are generated under torch when it is built with BUILD_SPLIT_CUDA=1 or BUILD_SPLIT_CUDA=1. (https://discuss.pytorch.org/t/no-libtorch-cuda-cpp-so-available-when-build-pytorch-from-source/159864). This link says that BUILD_SPLIT_CUDA is not default because
there may be other side effects (like increased binary size) that users might not be expecting, and it's only when we are compiling for many architectures where we run into these linker issues.
Currently, [torch](https://github.com/NixOS/nixpkgs/blob/nixos-22.11/pkgs/development/python-modules/torch/default.nix) doesn't have an option to it. Maybe an option like mklDnnSupport so that it can be turned on for packages like dgl-cuda116 that needs those files would be great. Any thought on this?
| 11:44:06 |
| @box1:matrix.org | * I'm trying to package dgl-cu116(dgl with cuda support) and it fails to find rpath for libtorch_cuda_cpp.so and libtorch_cuda_cu.so.
After some searching, those files are generated under torch when it is built with BUILD_SPLIT_CUDA=1 or BUILD_SPLIT_CUDA=1. (https://discuss.pytorch.org/t/no-libtorch-cuda-cpp-so-available-when-build-pytorch-from-source/159864). This link says that BUILD_SPLIT_CUDA is not default because
there may be other side effects (like increased binary size) that users might not be expecting, and it's only when we are compiling for many architectures where we run into these linker issues.
Currently, [torch](https://github.com/NixOS/nixpkgs/blob/nixos-22.11/pkgs/development/python-modules/torch/default.nix) doesn't have an option to it. Maybe an option like mklDnnSupport so that it can be turned on for packages like dgl-cuda116 that needs those files would be great. Any thought on this?
| 11:44:26 |
| @box1:matrix.org | * I'm trying to package dgl-cu116(dgl with cuda support) and it fails to find rpath for libtorch_cuda_cpp.so and libtorch_cuda_cu.so.
After some searching, those files are generated under torch when it is built with BUILD_SPLIT_CUDA=ON or BUILD_SPLIT_CUDA=1. (https://discuss.pytorch.org/t/no-libtorch-cuda-cpp-so-available-when-build-pytorch-from-source/159864). This link says that BUILD_SPLIT_CUDA is not default because
there may be other side effects (like increased binary size) that users might not be expecting, and it's only when we are compiling for many architectures where we run into these linker issues.
Currently, [torch](https://github.com/NixOS/nixpkgs/blob/nixos-22.11/pkgs/development/python-modules/torch/default.nix) doesn't have an option to it. Maybe an option like mklDnnSupport so that it can be turned on for packages like dgl-cuda116 that needs those files would be great. Any thought on this?
| 11:44:48 |
| @box1:matrix.org | * I'm trying to package dgl-cu116(dgl with cuda support) and it fails to find rpath for libtorch_cuda_cpp.so and libtorch_cuda_cu.so.
After some searching, those files are generated under torch when it is built with BUILD_SPLIT_CUDA=ON or BUILD_SPLIT_CUDA=1. (https://discuss.pytorch.org/t/no-libtorch-cuda-cpp-so-available-when-build-pytorch-from-source/159864). This link says that BUILD_SPLIT_CUDA is not default because
there may be other side effects (like increased binary size) that users might not be expecting, and it's only when we are compiling for many architectures where we run into these linker issues.
Currently, torch doesn't have an option to it. Maybe an option like mklDnnSupport so that it can be turned on for packages like dgl-cuda116 that needs those files would be great. Any thought on this?
| 11:45:26 |
|  hexa changed their display name from hexa to hexa (22.11 now). | 13:09:03 |
| hexa changed their display name from hexa (22.11 now) to hexa. | 14:38:53 |
 danielrf | Hi, I have some recent work that might be of interest to the Nix CUDA community: jetpack-nixos (https://github.com/anduril/jetpack-nixos)
See also this announcement post on the discourse: https://discourse.nixos.org/t/jetpack-nixos-nixos-module-for-nvidia-jetson-devices/23632 | 19:50:11 |
| danielrf |
The CUDA version included with jetpack is apparently not the same as just the aarch64 CUDA for servers, but I've tried to repackage the debs from NVIDIA in a way similar to cudaPackages in nixpkgs: https://github.com/anduril/jetpack-nixos/blob/master/cuda-packages.nix | 19:50:23 |
| 2 Dec 2022 |
| Samuel Ainsworth | yeah IIUC the best way would be to add an option to the pytorch derivation to enable BUILD_SPLIT_CUDA. Shouldn't be too hard to accomplish 🤞 | 01:29:56 |
| 4 Dec 2022 |
|  CRTified (old handle) changed their display name from CRTified to CRTified (old handle). | 14:19:44 |
| 6 Dec 2022 |
| CRTified (old handle) changed their profile picture. | 14:11:02 |
|  connor (he/him) left the room. | 15:17:44 |
| 12 Dec 2022 |
 SomeoneSerge (matrix works sometimes) | Just noticed that the GH action that triggers the Hercules rebuilds has been broken for 20 days :-( Updated it, but this only shows that it's something to be watched | 14:06:09 |
| 15 Dec 2022 |
| hexa | wondering if anyone here is motivated to get torchaudio source builds into nixpkgs | 00:56:52 |
| hexa | I've gotten pretty far, but I currently fail at linking libsox | 00:57:17 |
| hexa | https://github.com/NixOS/nixpkgs/pull/203136 | 00:57:18 |
| 21 Dec 2022 |
|  jdaly joined the room. | 17:33:43 |
| jdaly | Our project has a dependency on the MS Onnxruntime w/CUDA support. At present the, MS OnnxRuntime 1.13 release requires Cuda 11.6 and cuDNN 8.2.4, however, search the NixOS repo, we find 'pkgs.cudaPackages_11_6.cudatoolkit' but various other versions of cuDNN (i.e. 8.3.2) -- but no 8.2.4 package.
A few weeks back, I went looking for help on NixOS Discourse (See https://discourse.nixos.org/t/how-to-install-a-specific-version-of-cuda-and-cudnn/21725) -- where SergeK kindly pointed me to this forum.
We're complete newbies when it comes to NixOS, so we're looking for some help for get the cuDNN 8.2.4 package added -- or maybe its already there but we just haven't unearthed it yet!
Anyway, any tips or suggestions to help us get this figured out would be greatly appreciated. Thx! | 17:45:23 |
| jdaly | * Our project has a dependency on the MS Onnxruntime w/CUDA support. At present the, MS OnnxRuntime 1.13 release requires Cuda 11.6 and cuDNN 8.2.4, however, searching the NixOS repo, we find the 'pkgs.cudaPackages_11_6.cudatoolkit' and various other versions of cuDNN (e.g.. 8.3.2) -- but no 8.2.4 package.
A few weeks back, I went looking for help on NixOS Discourse (See https://discourse.nixos.org/t/how-to-install-a-specific-version-of-cuda-and-cudnn/21725) -- where SergeK kindly pointed me to this forum.
We're complete newbies when it comes to NixOS, so we're looking for some help for get the cuDNN 8.2.4 package added -- or maybe its already there but we just haven't unearthed it yet!
Anyway, any tips or suggestions to help us get this figured out would be greatly appreciated. Thx! | 17:53:59 |
| tpw_rules | can you build stuff using a checkout of nixpkgs? | 23:25:22 |
| 22 Dec 2022 |
|  Shiroki Satsuki joined the room. | 17:59:25 |
| Shiroki Satsuki | Hello, I'm using shell.nix with Arch Linux, and I'm trying to use python310Packages.pycuda in my Python project, but when I import pycuda.driver, Python complained about missing of libcuda.so.1 | 18:04:05 |
