train-pv-cnn-1.yml

experiment_name: train-pv-cnn-1 # basename, abs path to checkpoint model file will use output_dir and HPARAMS
output_dir: gs://pv_mapping/models # GCS location to save model checkpoints and logs
seed: 5

# data-related
epochs: 10
batch_size: 32
buffer_size: 320

# model-related
# loss_function: binary_crossentropy

# training related
early_stopping_patience: null # null or int