Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('json', {}), NamedSplit('test'): ('text', {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

PyTorch

PyTorch is a python package that provides two high-level features:

  • Tensor computation (like numpy) with strong GPU acceleration
  • Deep Neural Networks built on a tape-based autograd system

You can reuse your favorite python packages such as numpy, scipy and Cython to extend PyTorch when needed.

Contents of the PyTorch image

This container has the PyTorch framework installed and ready to use. The pytorch python module is installed as part of a Python 3.5 Conda environment in /opt/conda/envs/pytorch-py35. Both the compiled pytorch libraries and the Python 3.5 environment are included in $PATH. As a result, running python from the command line executes a Python 3.5 interpreter by default. /opt/pytorch contains the complete source of this version of PyTorch.

Running PyTorch

You can choose to use PyTorch as provided by NVIDIA, or you can choose to customize it. Run pytorch as you would any python program: run a python script, open interactive session in ipython or jupyter notebook. Start your scripts or interactive sessions with import torch

Customizing PyTorch

You can customize PyTorch one of two ways:

(1) Modify the version of the source code in this container and run your customized version, or (2) use docker build to add your customizations on top of this container if you want to add additional packages.

NVIDIA recommends option 2 for ease of migration to later versions of the PyTorch container image.

For more information, see https://docs.docker.com/engine/reference/builder/ for a syntax reference. Several example Dockerfiles are provided in the container image in /workspace/docker-examples.

Suggested Reading

For more information about pytorch, see

Downloads last month
56