TensorFlow Lite：初始化节点不存在

Question

我试图使用以下代码将Keras文件（.h5）中的模型转换为TensorFlow Lite文件（.tflite）：

# Save model as .h5 keras file
keras_file = "eSleep.h5"
model_save = tf.keras.models.save_model(model,keras_file,overwrite=True,include_optimizer=True)

# Export keras file to TensorFlow Lite model
converter = tf.lite.TFLiteConverter.from_keras_model_file(keras_file)
tflite_model = converter.convert()
open("eSleep.tflite", "wb").write(tflite_model)

但是，以下行：

tflite_model = converter.convert()

返回的错误：

I tensorflow/core/grappler/devices.cc:53] Number of eligible GPUs (core count >= 8): 0 (Note: TensorFlow was not compiled with CUDA support)
I tensorflow/core/grappler/clusters/single_machine.cc:359] Starting new session
E tensorflow/core/grappler/grappler_item_builder.cc:636] Init node dense/kernel/Assign doesn't exist in graph

有人可以帮助我了解“图形中不存在初始化节点密集/内核/分配”的含义是什么以及如何解决该错误？

Answer 1

我使用 TensorFlow 1.12 解决了这个问题。