| 16 Jul 2025 |
connor (burnt/out) (UTC-8) | See the end of the first section https://github.com/NixOS/nixpkgs/blob/master/doc/languages-frameworks/cuda.section.md#cuda-cuda | 07:02:27 |
| 18 Jul 2025 |
connor (burnt/out) (UTC-8) | Could I get a review on https://github.com/NixOS/nixpkgs/pull/426280? | 19:20:16 |
| 21 Jul 2025 |
connor (burnt/out) (UTC-8) | Went ahead and merged it | 17:20:26 |
| 23 Jul 2025 |
apyh | oof the nccl version in nixpkgs is quite old now | 16:30:38 |
apyh | (quite old in the ml world, lol. only a month old) | 16:31:25 |
apyh | torchtitan needs torch 2.8, torch 2.8 requires nccl 2.27, gotta update nccl myself | 16:31:49 |
apyh | guess I'll pr to nixpkgs lol | 16:31:56 |
apyh | pr opened 😁 | 16:59:39 |
Gaétan Lepage | Can you share the link apyh? | 22:56:02 |
apyh | ah sure! https://github.com/NixOS/nixpkgs/pull/427804 | 23:00:23 |
apyh | they added a bunch of new stuff so i have to patch the shebang in a second python script. surprisingly didn't cause a build failure without it, just didn't export some of the new symbols | 23:01:02 |
Gaétan Lepage | Thanks! | 23:03:49 |
| 24 Jul 2025 |
apyh | huh. thanks for the nixpkgs-review. very strange to me that it fails to build pytorch as a result, but that the python 3.13 failure is just a bunch of .. warnings inside torch? i'll compile again locally to see.. | 14:56:24 |
apyh | can't repro the build failure locally for python312Packages.torchWithCuda Gaétan Lepage 🤔 left a comment here to that effect https://github.com/NixOS/nixpkgs/pull/427804#issuecomment-3114819745 | 20:26:13 |
apyh | can't repro any of the build failures in fact, only took 3.5 hours per torch to test 😭 | 23:51:03 |
| 25 Jul 2025 |
Gaétan Lepage | It probably failed because of flakiness | 10:57:16 |
apyh | rebased it btw :) | 17:29:53 |
apyh | both builds worked fine on my machine.. does nixpkgs-review have a timeout? lol | 17:30:06 |
apyh | i have a 7800x3d and it still took 3.5 hours per torch build | 17:30:26 |
| 26 Jul 2025 |
Tristan Ross | Is that a PR that my 128 cores could be useful with? | 00:34:02 |
apyh | haha i mean, if you have the ram to match ;) | 01:07:29 |
apyh | it builds fine on my end - just a verification from someone else would be nice :) | 01:07:40 |
| oak 🏳️🌈♥️ changed their profile picture. | 08:29:04 |
Gaétan Lepage | Any objection to merging the nccl bump?
https://github.com/NixOS/nixpkgs/pull/427804 | 09:26:39 |
| 3 Aug 2025 |
connor (burnt/out) (UTC-8) | As a heads up, I’ll be removing CUDA versions prior to 11.8 from Nixpkgs and moving them to https://github.com/nix-community/cuda-legacy so they’ll be available as an overlay using the cool new functionality I got merged over the last few months to make it easier to extend CUDA stuff out of tree | 19:12:49 |
| 4 Aug 2025 |
aidalgol | I'm trying to write a derivation for https://github.com/alicevision/popsift/ and the configure phase fails because it cant find "cudadevrt".
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
CUDA_CUDADEVRT_LIBRARY
linked by target "popsift" in directory /build/source/src
linked by target "popsift-match" in directory /build/source/src/application
linked by target "popsift-demo" in directory /build/source/src/application
CUDA_cudadevrt_LIBRARY (ADVANCED)
linked by target "popsift" in directory /build/source/src
I haven't been able to find much about this beyond that it is the CUDA Device Runtime library. A grep for "devrt" on nixpkgs turns up nothing. Anyone here know about this CUDA library?
| 11:00:58 |
Robbie Buxton | It might be the dev output of cuda_cudart, I.e cuda_cudart.dev | 11:02:21 |
Robbie Buxton | I remember running into something similar in the past and I think I fixed it that way but might be misremembering | 11:03:16 |
aidalgol | Yay, now I'm on to compilation errors! | 11:05:39 |
aidalgol | Looks like this will only build on ooold CUDA versions. | 11:06:05 |