Google I/O is a wrap! Catch up on TensorFlow sessions View sessions

tflite_support.metadata.MetadataPopulator

Packs metadata and associated files into TensorFlow Lite model file.

MetadataPopulator can be used to populate metadata and model associated files into a model file or a model buffer (in bytearray). It can also help to inspect list of files that have been packed into the model or are supposed to be packed into the model.

The metadata file (or buffer) should be generated based on the metadata schema: third_party/tensorflow/lite/schema/metadata_schema.fbs

Example usage:

Populate matadata and label file into an image classifier model.

First, based on metadata_schema.fbs, generate the metadata for this image classifer model using Flatbuffers API. Attach the label file onto the ouput tensor (the tensor of probabilities) in the metadata.

Then, pack the metadata and label file into the model as follows.

  # Populating a metadata file (or a metadta buffer) and associated files to
  a model file:
  populator = MetadataPopulator.with_model_file(model_file)
  # For metadata buffer (bytearray read from the metadata file), use:
  # populator.load_metadata_buffer(metadata_buf)
  populator.load_metadata_file(metadata_file)
  populator.load_associated_files([label.txt])
  # For associated file buffer (bytearray read from the file), use:
  # populator.load_associated_file_buffers({"label.txt": b"file content"})
  populator.populate()

  # Populating a metadata file (or a metadta buffer) and associated files to
  a model buffer:
  populator = MetadataPopulator.with_model_buffer(model_buf)
  populator.load_metadata_file(metadata_file)
  populator.load_associated_files([label.txt])
  populator.populate()
  # Writing the updated model buffer into a file.
  updated_model_buf = populator.get_model_buffer()
  with open("updated_model.tflite", "wb") as f:
    f.write(updated_model_buf)

  # Transferring metadata and associated files from another TFLite model:
  populator = MetadataPopulator.with_model_buffer(model_buf)
  populator_dst.load_metadata_and_associated_files(src_model_buf)
  populator_dst.populate()
  updated_model_buf = populator.get_model_buffer()
  with open("updated_model.tflite", "wb") as f:
    f.write(updated_model_buf)

Note that existing metadata buffer (if applied) will be overridden by the new metadata buffer.

model_file valid path to a TensorFlow Lite model file.

IOError File not found.
ValueError the model does not have the expected flatbuffer identifer.

Methods

get_model_buffer

View source

Gets the buffer of the model with packed metadata and associated files.

Returns
Model buffer (in bytearray).

get_packed_associated_file_list

View source

Gets a list of associated files packed to the model file.

Returns
List of packed associated files.

get_recorded_associated_file_list

View source

Gets a list of associated files recorded in metadata of the model file.

Associated files may be attached to a model, a subgraph, or an input/output tensor.

Returns
List of recorded associated files.

load_associated_file_buffers

View source

Loads the associated file buffers (in bytearray) to be populated.

Args
associated_files a dictionary of associated file names and corresponding file buffers, such as {"file.txt": b"file content"}. If pass in file paths for the file name, only the basename will be populated.

load_associated_files

View source

Loads associated files that to be concatenated after the model file.

Args
associated_files list of file paths.

Raises
IOError File not found.

load_metadata_and_associated_files

View source

Loads the metadata and associated files from another model buffer.

Args
src_model_buf source model buffer (in bytearray) with metadata and associated files.

load_metadata_buffer

View source

Loads the metadata buffer (in bytearray) to be populated.

Args
metadata_buf metadata buffer (in bytearray) to be populated.

Raises
ValueError The metadata to be populated is empty.
ValueError The metadata does not have the expected flatbuffer identifer.
ValueError Cannot get minimum metadata parser version.
ValueError The number of SubgraphMetadata is not 1.
ValueError The number of input/output tensors does not match the number of input/output tensor metadata.

load_metadata_file

View source

Loads the metadata file to be populated.

Args
metadata_file path to the metadata file to be populated.

Raises
IOError File not found.
ValueError The metadata to be populated is empty.
ValueError The metadata does not have the expected flatbuffer identifer.
ValueError Cannot get minimum metadata parser version.
ValueError The number of SubgraphMetadata is not 1.
ValueError The number of input/output tensors does not match the number of input/output tensor metadata.

populate

View source

Populates loaded metadata and associated files into the model file.

with_model_buffer

View source

Creates a MetadataPopulator object that populates data to a model buffer.

Args
model_buf TensorFlow Lite model buffer in bytearray.

Returns
A MetadataPopulator(_MetadataPopulatorWithBuffer) object.

Raises
ValueError the model does not have the expected flatbuffer identifer.

with_model_file

View source

Creates a MetadataPopulator object that populates data to a model file.

Args
model_file valid path to a TensorFlow Lite model file.

Returns
MetadataPopulator object.

Raises
IOError File not found.
ValueError the model does not have the expected flatbuffer identifer.

METADATA_FIELD_NAME 'TFLITE_METADATA'
METADATA_FILE_IDENTIFIER b'M001'
TFLITE_FILE_IDENTIFIER b'TFL3'