An open-source framework for machine learning and other computations on decentralized data.
APACHE-2.0 License
Bot releases are hidden (Show)
Published by jkr26 over 1 year ago
pytype
dependency from TFF.tff.learning.algorithms.build_fed_recon_eval
now supports TFF optimizers.tff.types.deserialize_type
to not accept/return None
.tff.framework.ComputationBuildingBlock.is_foo
methods.tff.learning.algorithms.build_personalization_eval
totff.learning.algorithms.build_personalization_eval_computation
tff.learning.models.ReconstructionModel.from_keras_model
will now checktff.learning.models.ReconstructionModel.has_only_global_variables
Published by xiaoyux11 over 1 year ago
worker_binary
.FilteringReleaseManager
.build_personalization_eval
tobuild_personalization_eval_computation
.tff.to_type
implementation and type annotation to notNone
.program
package.tff.program.NativeFederatedContext
handles arguments ofPublished by nicolemitchell over 1 year ago
tff.learning.models.FunctionalModel
to allowPrefetchingDataSource
back to the tff.program
API now that thetff.simulation.compose_dataset_computation_with_learning_process
to be atff.learning.templates.LearningProcess
.prefetching_data_source_test
.tff.learning.optimizers
.Published by zcharles8 over 1 year ago
attrs
as containers in the tff.program
API.personalization_eval
module to the algorithms package.tff.learning.build_local_evaluation
API.tff.learning.reconstruction
to the tff.learning.algorithms
package.dm-tree
version 0.1.8
.dp-accounting
version 0.4.1
.tensorflow-privacy
version 0.8.9
.Published by michaelreneer over 1 year ago
tff.learning
.tff.learning.programs.EvaluationManager
to clean up statestff.learning.framework.ServerState
symbol.Published by jkr26 over 1 year ago
nest_asyncio
dependency from tutorials.tff.aggregators.DifferentiallyPrivateFactory.tree_adaptive
fortff.learning.programs.EvaluationManager
to set the evaluationtf.data.Dataset
iterator ops onto GPUs.Published by xiaoyux11 over 1 year ago
tff.learning.programs.EvaluationManager
, this enablesEvaluationManager
s from existing ones.attrs
in tff.aggregators
with typing.NamedTuple
.run_server
and server_context
from the tff.simulation
API.tff.framework
API:
tff.framework.local_executor_factory
tff.framework.DataBackend
tff.framework.DataExecutor
tff.framework.EagerTFExecutor
Published by huili0140 over 1 year ago
tff.learning.templates.LearningProcess
to allow non-sequence CLIENTS arguments.tff.simulation.compose_dataset_computation_with_learning_process
now returns a tff.learning.templates.LearningProcess
.tff.program.FederatedDataSourceIterator
s so that they can be serialized.forward_pass
attribute from the FunctionalModel
interface.from_keras_model
, MetricsFinalizersType
, BatchOutput
, Model
, and ModelWeights
symbols from the tff.learning
package. Users should instead use the tff.learning.models
package for these symbols.tff.learning.federated_aggregate_keras_metric
function.tff.simulation.compose_dataset_computation_with_learning_process
.tff.framework.remote_executor_factory_from_stubs
.tff.backends.xla
APIs.tff.backends.test
APIs to: tff.backends.test.(create|set)_(sync|async)_test_cpp_execution_context
.Published by ZacharyGarrett over 1 year ago
tff.backends.mapreduce.consolidate_and_extract_local_processing
astff.Serializable
attrs
classes as containers in the tff.program
API.tff.learning.algorithms
implementations to usetff.learning.models.FunctionalModel.loss
instead ofFunctionalModel.forward_pass
.sys.stdout
and sys.stderr
in subprocess.Popen
whenSequenceExecutor
to the C++ execution stack to handle sequence_*
Published by wushanshan over 1 year ago
loss_fn
to tff.learning.models.FunctionalModel
,type
field of Intrinsic
DTensor
based executor.tff.framework.DataBackend
. Python execution is deprecatednp.bytes_
types that incorrectly truncate byte stringPublished by zcharles8 over 1 year ago
tff.learning.algorithms.build_weighted_fed_avg_with_optimizer_schedule
tff.framework.local_executor_factory
tff.framework.remote_executor_factory_from_stubs
tff.framework.DataExecutor
tff.framework.EagerTFExecutor
tff.backends.native.create_local_python_execution_context
tff.framework.remote_executor_factory
executors_errors
module from the tff.framework
API, usetff.framework.RetryableError
instead.Published by michaelreneer over 1 year ago
tff.program.PrefetchingDataSource
, thetff.backends.native.create_local_python_execution_context
tff.backends.native.create_remote_python_execution_context
tff.backends.native.create_remote_async_python_execution_context
tff.backends.native.set_remote_async_python_execution_context
tff.backends.native.set_local_python_execution_context
tff.backends.native.set_remote_python_execution_context
tff.frameowrk.FederatingExecutor
tff.framework.ComposingExecutorFactory
tff.framework.ExecutorValue
tff.framework.Executor
tff.framework.FederatedComposingStrategy
tff.framework.FederatedResolvingStrategy
tff.framework.FederatingStrategy
tff.framework.ReconstructOnChangeExecutorFactory
tff.framework.ReferenceResolvingExecutor
tff.framework.RemoteExecutor
tff.framework.ResourceManagingExecutorFactory
tff.framework.ThreadDelegatingExecutor
tff.framework.TransformingExecutor
tff.framework.UnplacedExecutorFactory
tff.framework
, instead use:
tff.types.type_from_tensors
tff.types.type_to_tf_tensor_specs
tff.types.deserialize_type
tff.types.serialize_type
tff.learning.Model
to tff.learning.models.VariableModel
.cpp_execution_context.(create|set)_local_async_cpp_execution_context
execution_context.(create|set)_(sync|async)_local_cpp_execution_context
.Published by xiaoyux11 over 1 year ago
cpp_execution_context.(create|set)_local_cpp_execution_context
execution_context.(create|set)_(sync|async)_local_cpp_execution_context
.ExecutorService
from the public API.executors
, types
, and core
package.Published by wushanshan over 1 year ago
LayoutMap
message in the computation proto for TensorFlowDTensor
based execution.compiler_fn
parameter from the high level*_mergeable_execution_context
functions.tff.program.NativeFederatedContext
and thetff.program.PrefetchingDataSource
.build_functional_model_delta_update
to use ReduceDataset
ops toPublished by ZacharyGarrett over 1 year ago
tff.backends.native.desugar_and_transform_to_native
to the publicGroupNorm
implementation with implementation from Keras.tff.simulations.datasets.flair
APIs for the FLAIR dataset.model_output_manager
used intff.learning.programs
tff.learning.algorithms.build_weighted_fed_prox
parameterproximal_strength = 0.0
, matching the pydoc.Published by amlanchakoty almost 2 years ago
CppToPythonExecutorBridge
into the CPPExecutorFactory
.stream_structs
totff.backends.native.set_localhost_cpp_execution_context()
API.tff.backends.native.set_localhost_cpp_execution_context()
tobackends.native.set_sync_local_cpp_execution_context()
.tff.framework.ExecutionContext
totff.framework.SyncExecutionContext
to be consistent withtff.framework.AsyncExecutionContext
.SyncSerializeAndExecuteCPPContext
andAsyncSerializeAndExecuteCPPContext
classes.typing.Generic
in the learning package.Published by nicolemitchell almost 2 years ago
tff.program
API.BatchOutput
structure from the model forward_pass
(not just the predictions).tff.program.NativeFederatedContext
.Published by hbmcmahan almost 2 years ago
tff.learning.models.FunctionalModel
tff.programs.FileProgramStateManager
copy.deepcopy
for structures of awaitables (non-pickable)tff.learning.programs
.tff.learning.programs.EvaluationManager
where the restarted evaluationcom_google_protobuf
version to v3.19.0
.six
.tff.program.ReleaseManager
.tff.simulation.build_uniform_sampling_fn
so that the outputPublished by jkr26 almost 2 years ago
tff.learning.programs
for federated program-logic using thetff.program
APIs.tensorflow
to version 2.11.0
.tensorflow_compression
to version 2.11.0
.bazel_skylib
to version 1.3.0
.