I’ve been exploring the challenges of deploying PyTorch models on resource-constrained microcontrollers, and it’s apparent that PyTorch lacks native support for such environments. This has led me to explore external tools like TVM.
While I see most tutorials with TFlite integration, I’m eager to understand if there are plans or ongoing efforts to expand support for PyTorch on microcontrollers using TVM. I’ve noticed there’s only one (limited) tutorial available.
Any insights or resources you can share regarding PyTorch and TVM integration for edge deployment would be greatly appreciated. I’d love to hear what you guys recommend as the best approach for integrating PyTorch with TVM micro.