Known Issues

  • TorchQueueDataset’s multiprocessing does not work correctly if other multiprocessing related functions are called before initializing the Queue. This currently also prevents us from using two Queues simultaneously.

  • TorchQueueDataset will likely not work correctly with Distributed Training. Could not test yet.

Frequently Asked Questions

Go ask us some questions on GitHub!