Hi,
Thanks for sharing this implementation. This is quite helpful for detailed study.
I tried running it for 13B parameters but I am getting errored out saying trying to take out more parameters than is in the tensor.
I am using weights downloaded from https://github.com/facebookresearch/metaseq/tree/main/projects/OPT as suggested in the repo.
Is anyone else facing the issue?
Thanks,
Deeksha
Hi,
Thanks for sharing this implementation. This is quite helpful for detailed study.
I tried running it for 13B parameters but I am getting errored out saying trying to take out more parameters than is in the tensor.
I am using weights downloaded from https://github.com/facebookresearch/metaseq/tree/main/projects/OPT as suggested in the repo.
Is anyone else facing the issue?
Thanks,
Deeksha