Error during conversion from saved model to tflite

TensorFlow General Discussion
, , ,
1

My command is as follows.

tflite_convert --saved_model_dir='/home/export/best_model_d0/1/' --graph_def_file='/home/u291321/Python_work/video_compression/src/git_tf2_effdet/EfficientDet-tensorflow2-main/export/best_model_d0/1/saved_model.pb' --output_file='model_clpb.tflite' --input_shapes=1,128,80,1 --input_arrays='input' --output_arrays='Softmax'

My error is,

raise ConverterError(str(e))
tensorflow.lite.python.convert.ConverterError: <unknown>:0: error: loc("Func/StatefulPartitionedCall/input/_0"): requires all operands and results to have compatible element types
<unknown>:0: note: loc("Func/StatefulPartitionedCall/input/_0"): see current operation: %1 = "tf.Identity"(%arg0) {_class = ["loc:@Func/StatefulPartitionedCall/functional_1/lambda/StatefulPartitionedCall/input/_715"], device = ""} : (tensor<?x?x?x3x!tf.quint8>) -> tensor<?x?x?x3xui8>

Please help in resolving this.

3

Hi @Dibya_Jyoti, I have tried to convert a sample model to tflite using the following command

!tflite_convert --saved_model_dir='/content/model' 
                --graph_def_file='/content/model/saved_model.pb' 
                --output_file='model1_clpb.tflite' --input_shapes=1,28,28,1 
                --input_arrays='input' --output_arrays='Softmax'

and did not face any error. please refer to this gist for working code example.

Could you please share the model architecture which you are trying to convert. Thank You.