You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I would be eager to port myself the operation from lite to micro following this guide, however, the guide seems outdated and, as I mentioned before, the kernel for the lite list is missing.
Any plans on including back such an operation? Is there anything we can do to help?
What is the recommended approach to port bidirectional RNNs to TFLM if the ReverseV2 op. cannot be used?
The text was updated successfully, but these errors were encountered:
ReverseV2 is a TF operator used for bidirectional RNNs. I am in need of using it in TFLM.
Back in 2019 it was added to the repo in this commit: rkuester/tensorflow@d5ee710 and today some traces of it can be found in
tflite-micro/tensorflow/lite/builtin_ops.h
Line 135 in ef64591
However, the kernel from such op is missing from the list of lite kernels: https://github.com/tensorflow/tflite-micro/tree/main/tensorflow/lite/kernels/internal/reference
I would be eager to port myself the operation from lite to micro following this guide, however, the guide seems outdated and, as I mentioned before, the kernel for the lite list is missing.
Any plans on including back such an operation? Is there anything we can do to help?
What is the recommended approach to port bidirectional RNNs to TFLM if the ReverseV2 op. cannot be used?
The text was updated successfully, but these errors were encountered: