Exporting model using Tf.saved_model

Deependra_Parichha · May 11, 2022, 11:23am

What happens(under the hood) after exporting the model in tf.saved_model?

Bhack · May 11, 2022, 11:29am

tf.saved_model.save:

Exports a tf.Module (and subclasses) obj to SavedModel format.

You can see impl details at:

tensorflow/tensorflow/blob/v2.8.0/tensorflow/python/saved_model/save.py#L1152-L1335


      
          @tf_export(
              "saved_model.save",
              v1=["saved_model.save", "saved_model.experimental.save"])
          def save(obj, export_dir, signatures=None, options=None):
            # pylint: disable=line-too-long
            """Exports a [tf.Module](https://www.tensorflow.org/api_docs/python/tf/Module) (and subclasses) `obj` to [SavedModel format](https://www.tensorflow.org/guide/saved_model#the_savedmodel_format_on_disk).
          
            The `obj` must inherit from the [`Trackable` class](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/training/tracking/base.py#L591).
          
            Example usage:
          
            >>> class Adder(tf.Module):
            ...   @tf.function(input_signature=[tf.TensorSpec(shape=[], dtype=tf.float32)])
            ...   def add(self, x):
            ...     return x + x
          
            >>> model = Adder()
            >>> tf.saved_model.save(model, '/tmp/adder')
          
            The resulting SavedModel is then servable with an input named "x", a scalar

This file has been truncated. show original