AM69A: Running TF Lite Model on C7x within AM69A

TSRDM

Part Number: AM69A

Tool/software:

Hello e2e experts, our customer is seeking advise on how to deploy a custom TFLite floating point model for Audio AI algorithm on AM69A's C7x DSP core in Floating Point with no MMA. If it can be done on cloud connected EVMs that is the best, we’d need a guidance how to do it.

Step by Step instructions are greatly appreciated.

Thank you!

1 month ago

0 Wen Li 1 month ago

TI__Expert 4471 points

Hello;

1. For step by step introduction instructions; the jupyter notebook examples are easy to follow, great examples. The following is the link.

https://github.com/TexasInstruments/edgeai-tidl-tools/tree/master/examples/jupyter_notebooks

2. For python script based examples and production example. You can follow the examples

https://github.com/TexasInstruments/edgeai-tidl-tools/tree/master/examples/osrt_python

There are examples for three formats: ONNX, TFL, and TVM. The "tflrt_delegate.py" is the TFLite runtime example.

During the learning process, if you have questions, please feel free to submit the new questions.

Thanks and regards

Wen Li

Processors

Processors forum

AM69A: Running TF Lite Model on C7x within AM69A