我正在尝试将bert服务模型转换为其等效的public class Word
{
public string word;
private int typeIndex; // Checks for current letter
private int wordIndex; // index of word used from WordGenerator
WordDisplay display;
public Word(WordDisplay _display) // Displays the word as Hiragana / Katakana
{
wordIndex = WordGenerator.GetIndex(); // now it will only be called ONCE per instance
word = _WordGenerator.wordList_Romaji[wordIndex];
// you can get the equivalent? letters similarly...with something like:
word = _WordGenerator.wordList_Hiragana[wordIndex];
display = _display;
display.SetWord(word);
}
// ... other existing code ...
}
格式。
用于转换为tflite格式的代码段为：
tflite如果输入形状为import os
import tensorflow as tf
cur_dir = os.getcwd()
saved_model_dir = os.path.join(cur_dir, 'serving_dir')
converter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)
converter.target_ops = [tf.lite.OpsSet.TFLITE_BUILTINS, tf.lite.OpsSet.SELECT_TF_OPS]#,
#tf.lite.OpsSet.TFLITE_BUILTINS]
tflite_model = converter.convert()
open(os.path.join(cur_dir, 'cnv1.tflite'), 'wb').write(tflite_model)
，则出现以下错误：
None批处理大小已在代码中指定。如果我将形状定为Traceback (most recent call last):
File "temp.py", line 13, in <module>
tflite_model = converter.convert()
File "/media/data2/anaconda3/envs/tf-lite/lib/python3.6/site-packages/tensorflow/lite/python/lite.py", line 406, in convert
"'{0}'.".format(_tensor_name(tensor)))
ValueError: Provide an input shape for input array 'input_example_tensor'.
，则可以转换为等效的(10,)。为了对生成的模型（tflite进行推断，已使用here中的代码段。但是tflite抛出以下内容：
interpreter.allocate_tensors()有人在github上报告了错误。为了解决这个问题，jdduke发表了评论：
我们希望在1.14版本中解决该问题。
我已经尝试Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/media/data2/anaconda3/envs/tf-lite/lib/python3.6/site-packages/tensorflow/lite/python/interpreter.py", line 73, in allocate_tensors
return self._interpreter.AllocateTensors()
File "/media/data2/anaconda3/envs/tf-lite/lib/python3.6/site-packages/tensorflow/lite/python/interpreter_wrapper/tensorflow_wrap_interpreter_wrapper
.py", line 106, in AllocateTensors
return _tensorflow_wrap_interpreter_wrapper.InterpreterWrapper_AllocateTensors(self)
RuntimeError: Regular TensorFlow ops are not supported by this interpreter. Make sure you invoke the Flex delegate before inference.Node number 0 (Fle
x) failed to prepare.
来获取以下内容：
1.14.0rc0