How can we run this on multiple GPUs?

[elena-nuntio]

I've been trying to run this code on a multiple GPUs machine, by providing multiple GPU indexes in the CUDA_VISIBLE_DEVICES variable, but my machine always ends up running it on only one. Has anyone had success with this?

[JackLangerman]

I have also been working on this with not much success. Please let me know if you find a solution. I'll do the same. I also tried wrapping the model in Torch.DataParallel but no luck.