Some of the operators in the model are not supported by the standard TensorFlow Lite runtime.

心已赠人 2022-12-16 13:07 167阅读 0赞

用TensorFlow Lite转换模型时，报如下错误：

Some of the operators in the model are not supported by the standard TensorFlow Lite 
    runtime. If those are native TensorFlow operators, you might be able to use the extended 
    runtime by passing --enable_select_tf_ops, or by setting 
    target_ops=TFLITE_BUILTINS,SELECT_TF_OPS when calling tf.lite.TFLiteConverter(). 
    Otherwise, if you have a custom implementation for them you can disable this error with --
    allow_custom_ops, or by setting allow_custom_ops=True when calling 
    tf.lite.TFLiteConverter(). Here is a list of builtin operators you are using: ADD, 
    AVERAGE_POOL_2D, CAST, CONCATENATION, CONV_2D, DEPTHWISE_CONV_2D, 
    FULLY_CONNECTED, GREATER_EQUAL, LESS, MUL, RESHAPE, SELECT, 
    TRANSPOSE. Here is a list of operators for which you will need custom implementations: 
    RandomUniform.

解决方法：

import tensorflow as tf
    
    converter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)
    # 在convert()前加上这句
    converter.allow_custom_ops=True
    
    tflite_model = converter.convert()
    open("converted_model.tflite", "wb").write(tflite_model)

参考：[Some of the operators in the model are not supported by the standard TensorFlow Lite runtime. If th][]

[Some of the operators in the model are not supported by the standard TensorFlow Lite runtime. If th]: https://www.it610.com/article/1281539546202259456.htm