rknn-toolkit的tflite转换接口是否不支持Lambda？

我使用rknn.load_tflite接口进行模型的转换，报错
RKNN output shape(reshape): (1 186624)
D Tensor @model_1/model_1/flatten/Reshape_17ut0 type: float32
D Process model_1/model_1/dense/Relu_18 ...
D RKNN output shape(fullconnect): (1 128)
D Tensor @model_1/model_1/dense/Relu_18ut0 type: float32
D Process model_1/model_1/dense/Relu_19 ...
D RKNN output shape(relu): (1 128)
D Tensor @model_1/model_1/dense/Relu_19ut0 type: float32
D Process model_1/lambda/sub_20 ...
E get error shape: [None]
E Catch exception when loading tflite model: ./siamese.tflite!
E Traceback (most recent call last):
E   File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 437, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E   File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 437, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E   File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 437, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E   File "rknn/base/RKNNlib/RKNNnetbuilder.py", line 451, in rknn.base.RKNNlib.RKNNnetbuilder.RKNNNetBuilder.build_layer
E   File "rknn/api/rknn_log.py", line 323, in rknn.api.rknn_log.RKNNLog.e
E ValueError: get error shape: [None]
E Please feedback the detailed log file <log_feedback_to_the_rknn_toolkit_dev_team.log> to the RKNN Toolkit development team.


看起来是在lambda层转换出错，把改成去掉之后发现能正常转换，请问是不支持lambda吗？