I am working with tensorflow similarity. Even though i have converted models to tflite for deployment purposes before. I am not sure if it possible to include post processing steps to tflite or anything else for production.
What i want: I have trained a model. It is important to do indexing on data for inference before passing the test image/data. So i am wondering if it is possible to include those indexing steps into tflite or anyother way for production purposes??
# These are the steps necessary to produce results
# Thus i need a way to include this information for deployment reasons.
x_index, y_index = sampler.get_slice(begin=0, size=300)
index_data = tf.cast(x_index, dtype="int32") # casted so it can displayed
test_y = [int(c) for c in y_test]
model.reset_index()
model.index(x_index, y_index, data=x_index)
Many Thanks.
