
############################## Experiment details ##############################

dataset: speech_commands
dataset_dir: ../data/speech_commands/
output_dir: output/fedavg-speech_commands-2022-08-15_22:08:04/
model: m5
num_rounds: 30
eval_every: 10
ServerType: <class 'rayleaf.entities.server.Server'>
client_types: [(<class 'rayleaf.entities.client.Client'>, 20)]
clients_per_round: 20
client_lr: 0.05
batch_size: 64
seed: 0
use_val_set: False
num_epochs: 10
gpus_per_client_cluster: 1
num_client_clusters: 4
save_model: False

############################## Simulation ##############################

Spawning 4 ClientClusters using cuda device (this may take a while)
20 total clients: 20 Clients

>>> Training Accuracy: Round 0 <<<
+-----------+-----------+-------------------+-------------------+
|   average |    median |   10th percentile |   90th percentile |
|-----------+-----------+-------------------+-------------------|
| 0.0367268 | 0.0357599 |         0.0247599 |         0.0490132 |
+-----------+-----------+-------------------+-------------------+
>>> Test Accuracy: Round 0 <<<
+-----------+-----------+-------------------+-------------------+
|   average |    median |   10th percentile |   90th percentile |
|-----------+-----------+-------------------+-------------------|
| 0.0534351 | 0.0564388 |         0.0269559 |         0.0853448 |
+-----------+-----------+-------------------+-------------------+
--- Round 1 of 30: Training 20 clients: 20 Clients ---
