Giters

lucidrains / DALLE-pytorch

Implementation / replication of DALL-E, OpenAI's Text to Image Transformer, in Pytorch

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CUDA Error: unknown error when generating an image (model trained with --taming flag)

vladislabv opened this issue · comments

commented

Hello everyone!

I am facing an issue while trying to generate an image, after a model was successfully trained. The training have done with a docker image provided in the repository. After training, I also get True by testing torch.cuda in the python compiler, so I do not get where the error comes from.
Here is the error traceback:
cuda_error
Behind docker I am using Windows 10, Nvidia GeForce 1050 as GPU, Intel(R) Core(TM) i5-8300H CPU @ 2.30GHz 2.30 GHz as CPU and 8 GB RAM.
Here is the output of the cuda being used on docker:
nvcc_output
As well as drivers being used on the PC:
nvidia_smi_info

I would appreciate any help, if someone have experienced the same problem!