Tensorflow can't parse serialized Example

Question

I have the following dictionary:

{'points': <tf.Tensor 'IteratorGetNext:3' shape=(?, 50000, 7) dtype=float32>, 'point_indices': <tf.Tensor 'IteratorGetNext:2' shape=(?, 50000) dtype=int32>, 'image_path': <tf.Tensor 'IteratorGetNext:1' shape=(?,) dtype=string>, 'flags': <tf.Tensor 'IteratorGetNext:0' shape=(?, 2) dtype=int64>}

Once i run sess.run() and i pass the dictionary as parameter i get the following error:

tensorflow.python.framework.errors_impl.InvalidArgumentError: {{function_node __inference_Dataset_map_<class 'functools.partial'>_107}} Key: point_indices.  Can't parse serialized Example.
     [[{{node ParseSingleExample/ParseSingleExample}}]]
     [[IteratorGetNext]]

Any idea on how to solve this??