How to run tflite interpreter with custom ops of tensorflow_text ops in java android

19 views Asked by At

I am trying to use fastberttokenizer from tensorflow_text library which has tflite support to convert to tflite.I follow text this guide in tensorflow for conversion. There inference of tensorflow_text ops are given in python .The code in python is :

# Perform TensorFlow Lite inference.
interp = interpreter.InterpreterWithCustomOps(
    model_content=tflite_model,
    custom_op_registerers=tf_text.tflite_registrar.SELECT_TFTEXT_OPS)
interp.get_signature_list()

I want to acheive this inference for android java for tflite model .i have already successfully converted tf.text.Fastberttokenizer to tflite model .plz Help to Acheive this.

I succesfully converted tensorflow_text ops of Fastberttokenizer to tflite .i want to know how to infer this in android java with interpreter as Tensorflow office guid does not have snippet for java android

0

There are 0 answers