Skip to content

Known Issues and Limitations

  • nav.Module moves original torch.nn.Module to the CPU, in case of weight sharing that might result in unexpected behavior
  • For data dependent dynamic control flow (multiple computation graphs) nav.Module might copy the weights for each separate graph
  • Source model running in Python can cause OOM issue when GPU memory is larger than CPU RAM memory
  • Verify command could potentially experience CUDA OOM errors while trying to run inference on two models at the same time.
  • Dependencies between modules in optimized pipelines may lead to unexpected behavior and failure in Inplace Optimize