我正在嘗試將 TF 模型轉(zhuǎn)換為 TFLite。模型以 .pb格式保存，我使用以下代碼對其進行了轉(zhuǎn)換：import osimport tensorflow as tffrom tensorflow.core.protobuf import meta_graph_pb2export_dir = os.path.join('export_dir', '0')if not os.path.exists('export_dir'):    os.mkdir('export_dir')tf.compat.v1.enable_control_flow_v2()tf.compat.v1.enable_v2_tensorshape()# I took this function from a tutorial on the TF websitedef wrap_frozen_graph(graph_def, inputs, outputs):    def _imports_graph_def():        tf.compat.v1.import_graph_def(graph_def, name="")    wrapped_import = tf.compat.v1.wrap_function(_imports_graph_def, [])    import_graph = wrapped_import.graph    return wrapped_import.prune(            inputs, outputs)graph_def = tf.compat.v1.GraphDef()loaded = graph_def.ParseFromString(open(os.path.join(export_dir, 'saved_model.pb'),'rb').read())concrete_func = wrap_frozen_graph(        graph_def, inputs=['extern_data/placeholders/data/data:0', 'extern_data/placeholders/data/data_dim0_size:0'],    outputs=['output/output_batch_major:0'])concrete_func.inputs[0].set_shape([None, 50])concrete_func.inputs[1].set_shape([None])concrete_func.outputs[0].set_shape([None, 100])converter = tf.lite.TFLiteConverter.from_concrete_functions([concrete_func])converter.experimental_new_converter = Trueconverter.post_training_quantize=Trueconverter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS,                                               tf.lite.OpsSet.SELECT_TF_OPS]converter.allow_custom_ops=Truetflite_model = converter.convert()# Save the model.if not os.path.exists('tflite'):    os.mkdir('tflite')output_model = os.path.join('tflite', 'model.tflite')with open(output_model, 'wb') as f:     f.write(tflite_model)現(xiàn)在，我在代碼中找不到任何內(nèi)容VarHandleOp，我發(fā)現(xiàn)它實際上是在tensorflow中（https://www.tensorflow.org/api_docs/python/tf/raw_ops/VarHandleOp）。那么，為什么TFLite無法識別呢？