README.md

Hello PyTorch with Tensorboard Streaming

Example of using NVIDIA FLARE to train an image classifier using federated averaging (FedAvg) and PyTorch as the deep learning training framework.

This example also highlights the TensorBoard streaming capability from the clients to the server.

NOTE: This example uses the CIFAR-10 dataset and will load its data within the trainer code.

1. Install requirements and configure PYTHONPATH

Install additional requirements:

python -m pip install -r requirements.txt

Set PYTHONPATH to include custom files of this example:

export PYTHONPATH=${PWD}/..

2. Run the experiment

Use nvflare simulator to run the example:

nvflare simulator -w /tmp/nvflare/ -n 2 -t 2 ./jobs/tensorboard-streaming

3. Access the logs and results

You can find the running logs and results inside the simulator's workspace/simulate_job

$ ls /tmp/nvflare/simulate_job/
app_server  app_site-1  app_site-2  log.txt tb_events

4. Tensorboard Streaming

On the client side, TBWriter works as a TensorBoard SummaryWriter. Instead of writing to TB files, it actually generates NVFLARE events of type analytix_log_stats. The ConvertToFedEvent widget will turn the event analytix_log_stats into a fed event fed.analytix_log_stats, which will be delivered to the server side.

On the server side, the TBAnalyticsReceiver is configured to process fed.analytix_log_stats events, which writes received TB data into appropriate TB files on the server.

To view training metrics that are being streamed to the server, run:

tensorboard --logdir=/tmp/nvflare/simulate_job/tb_events

Note: If the server is running on a remote machine, use port forwarding to view the TensorBoard dashboard in a browser. For example:

ssh -L {local_machine_port}:127.0.0.1:6006 user@server_ip

NOTE: For a more in-depth guide about the TensorBoard streaming feature, see PyTorch with TensorBoard.