openfl.pipelines.tensor_codec.TensorCodec

class openfl.pipelines.tensor_codec.TensorCodec(compression_pipeline)

Bases: object

TensorCodec is responsible for the following.

1. Tracking the compression/decompression related dependencies of a given tensor. 2. Acting as a TensorKey aware wrapper for the compression_pipeline functionality.

Class Attributes:

compression_pipeline – The pipeline used for compression.
lossless_pipeline – The pipeline used for lossless compression.

Methods

`apply_delta`	Add delta to the nparray.
`compress`	Function-wrapper around the tensor_pipeline.forward function.
`decompress`	Function-wrapper around the tensor_pipeline.backward function.
`find_dependencies`	Resolve the tensors required to do the specified operation.
`generate_delta`	Create delta from the updated layer and base layer.
`set_lossless_pipeline`	Set lossless pipeline.

static apply_delta(tensor_key, delta, base_model_nparray, creates_model=False)

Add delta to the nparray.

Parameters:

tensor_key – This is the tensor_key associated with the delta. Should have a tag of ‘trained’ or ‘aggregated’.
delta – Weight delta between the new model and old model.
base_model_nparray – The nparray that corresponds to the prior weights.
creates_model – If flag is set, the tensorkey returned will correspond to the aggregator model.

Returns:

new_model_tensor_key – Latest model layer tensorkey.
new_model_nparray – Latest layer weights.

compress(tensor_key, data, require_lossless=False, **kwargs)

Function-wrapper around the tensor_pipeline.forward function.

It also keeps track of the tensorkeys associated with the compressed nparray.

Parameters:

tensor_key – TensorKey is provided to verify it should be compressed, and new TensorKeys returned will be derivatives of the existing tensor_name.
data – (uncompressed) numpy array associated with the tensor_key.
require_lossless – boolean. Does tensor require compression.

Returns:

compressed_tensor_key – Tensorkey corresponding to the decompressed tensor.
compressed_nparray – The compressed tensor.
metadata – metadata associated with compressed tensor.

decompress(tensor_key, data, transformer_metadata, require_lossless=False, **kwargs)

Function-wrapper around the tensor_pipeline.backward function.

It also keeps track of the tensorkeys associated with the decompressed nparray.

Parameters:

tensor_key – TensorKey is provided to verify it should be decompressed, and new TensorKeys returned will be derivatives of the existing tensor_name.
data – (compressed) numpy array associated with the tensor_key.
transformer_metadata – metadata associated with the compressed tensor.
require_lossless – boolean, does data require lossless decompression.

Returns:

decompressed_tensor_key – Tensorkey corresponding to the decompressed tensor.
decompressed_nparray – The decompressed tensor.

find_dependencies(tensor_key, send_model_deltas)

Resolve the tensors required to do the specified operation.

Parameters:

tensor_key – A tuple containing the tensor name, origin, round number, report, and tags.
send_model_deltas – A boolean flag indicating whether to send model deltas.

Returns:

tensor_key_dependencies – A list of tensor keys that are dependencies of the given tensor key.

static generate_delta(tensor_key, nparray, base_model_nparray)

Create delta from the updated layer and base layer.

Parameters:

tensor_key – This is the tensor_key associated with the nparray. Should have a tag of ‘trained’ or ‘aggregated’
nparray – The nparray that corresponds to the tensorkey.
base_model_nparray – The base model tensor that will be subtracted from the new weights.

Returns:

delta_tensor_key – Tensorkey that corresponds to the delta weight array.
delta – Difference between the provided tensors.

set_lossless_pipeline(lossless_pipeline)

Set lossless pipeline.

Parameters:: lossless_pipeline – The pipeline to be set as the lossless pipeline. It should be a pipeline that is not lossy.
Raises:: AssertionError – If the provided pipeline is not lossless.