Tim Dettmers

The binary that is used is determined at runtime. This means in your case there are two modes of failures: - the CUDA driver is not detected (`libcuda.so`) - the...

I believe this is fixed in the latest version. It prints instructions on how to debug the situation and alternatively prints out compilation instructions which should fix the issue.

I believe this is fixed in the latest version. It prints instructions on how to debug the situation and alternatively prints out compilation instructions which should fix the issue.

I am closing this as this issue is related to a part of the model being on the CPU which is currently managed by the accelerate library. If this is...

This is a great catch. Can you please submit this to the transformers github repo? This is only indirectly a bitsandbytes issue.

I looked at the deepspeed implementation before, which had a similar issue with shared weights. The problem was that the algorithm splits all tensors found in the optimizer state, which...

Amazing! Thank you for bringing this to my attention. I will try to get in touch with the author of the ROCm library and support AMD GPUs by default.

Sorry for taking so long on this. I am currently onboarding more maintainers and we should see some progress on this very soon. This is one of our high-priority issues.

Did this work out for you? It seems a straightforward fix and a good contribution if this would make the library jetson compatible.

Thank you so much for this contribution. I am sorry that it took so long to reply and look at this. We are currently working on integrating different devices other...