Text-to-Video
English

Colab not working

#1
by MicahB - opened

Hi there, I get the following output and error when attempting to run the colab you have linked on the model card. Would you mind taking a look at it and fixing the error if possible? Thanks.

WARNING:jax._src.xla_bridge:An NVIDIA GPU may be present on this machine, but a CUDA-enabled jaxlib is not installed. Falling back to cpu.
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-2-a6d6d9c9388e> in <cell line: 7>()
      5 from utils.unet import UNet3DConditionModel
      6 from utils.pipeline_magictime import MagicTimePipeline
----> 7 from utils.util import save_videos_grid
      8 from utils.util import load_weights
      9 

4 frames
/usr/local/lib/python3.10/dist-packages/torch/_library/abstract_impl.py in register(self, func, source)
     28                 f"{self.kernel.source}."
     29             )
---> 30         if torch._C._dispatch_has_kernel_for_dispatch_key(self.qualname, "Meta"):
     31             raise RuntimeError(
     32                 f"impl_abstract(...): the operator {self.qualname} "

RuntimeError: operator torchvision::nms does not exist

Thank you for your interest. It seems that the version of jax does not match the versions of cuda and cudnn. As this colab is not released by us, perhaps you can seek help from camenduru.

Sign up or log in to comment