I’m wondering if there’s an equivalent of TensorFlow.js in PyTorch?
TensorFlow.js works right in the browser and I’d like to do something similar with a PyTorch model (if possible).
Has anyone done this?
Or is using an API for deployment too easy these days?
Running directly in the browser has a fair few benefits.
There doesn’t seem overwhelming community demand for it but if you think I’m wrong please let me know!
An alternative may be https://github.com/microsoft/onnxjs but the repo is read only so idk if this the best option long term
You can do PyTorch → ONNX → ONNX Runtime Web.
ONNX Runtime Web is under active development.
FYI, ONNX.js has been replaced by ONNX Runtime Web which offers enhanced user experience and improved performance.