Plato v0.4.1

Fixed several important issues related to client-side samplers, loading custom algorithms, federated unlearning, and added default values for configurations.

Plato v0.4.0

Supported running an FL session on multiple GPUs, and further improved scalability in memory usage by always launching a constant number of client processes regardless of the number of clients selected per round. Made client simulation mode the default and only mode of operation.

Plato v0.3.9

Fixed an urgent issue related to loading custom trainers from examples/.

Plato v0.3.8

Miscellaneous improvements and bug fixes.

Plato v0.3.7

Added support for HuggingFace Language Modelling models and datasets, reinforcement learning servers, simulating client/server communication, measuring communication time, additional examples using the asynchronous mode, and removed wandb usage.

Plato v0.3.6

Bug fixes, added the EMNIST dataset and supported resuming a FL training session.

Plato v0.3.5

This release fixed several issues in async mode operation when the wall-clock time is simulated on the server.

Plato v0.3.4

Added several multi-modal data sources, and supported simulating the wall clock time in asynchronous mode, when the clients on the same physical machine are training in small batches (controlled by trainer -> max_concurrency) due to insufficient GPU memory.

Plato v0.3.3

Added support for differentially private training on the client side, fixed issues related to cross-silo training, and added basic support for asynchronous training with bounded staleness.

Plato v0.3.2

Added basic model and feature processors before data payloads are transmitted over the network.