tutorials
tutorials copied to clipboard
Tutorial for converting Onnx back to pytorch/tf trainnable graph/code?
I know people mostly use ONNX for compilation and exporting, but just wondering, is there a tool where people have tried to convert ONNX model back to trainable pytorch codes?
This is a very useful feature for cases where I preload a STOA model from ONNX and want to do some fine-tuning on it.
I could write a tool myself but just wondering if that has been done before.
If you are using MXNet, please have a look at https://mxnet.apache.org/api/python/docs/tutorials/packages/onnx/fine_tuning_gluon.html
@GritBear were you able to find any other resources on this?
did you guys find anything yet? i am looking for such a tool because it is theoretically possible to do