I'm having hard time inferencing with my custom trained tflite model. default ssd mobilenet model's output is TFLite_Detection_PostProcess but my custom model is StatefulPartitionedCall as below
so I tried
command = "tflite_convert \
--saved_model_dir={} \
--output_file={} \
--input_arrays=normalized_input_image_tensor \
--output_arrays='TFLite_Detection_PostProcess','TFLite_Detection_PostProcess:1','TFLite_Detection_PostProcess:2','TFLite_Detection_PostProcess:3' \
--inference_type=FLOAT \
--allow_custom_ops".format('tflite/saved_model', 'tflite/model88.tflite', )
but with model88.tflite which I think I wrote output_arrays with TFLite_Detection_PostProcess I searched with neutron to see output arrays but , it keep saying my tflite model output is still StatefulPartitionedCall