super-gradients/src/super_gradients/recipes/imagenet_efficientnet.yaml

# Efficientnet-B0 Imagenet training
# This example trains with effective batch size = 64 * 4 gpus = 256.
# Epoch time on 4 X 3090Ti distributed training is ~ 16:25 minutes
# Logs and tensorboards: s3://deci-pretrained-models/efficientnet_b0/
# Instructions:
#   0. Make sure that the data is stored in dataset_params.dataset_dir or add "dataset_params.data_dir=<PATH-TO-DATASET>" at the end of the command below (feel free to check ReadMe)
#   1. Move to the project root (where you will find the ReadMe and src folder)
#   2. Run the command:
#       python src/super_gradients/examples/train_from_recipe_example/train_from_recipe.py --config-name=imagenet_efficientnet

defaults:
  - training_hyperparams: imagenet_efficientnet_train_params
  - dataset_params: imagenet_efficientnet_dataset_params
  - arch_params: efficientnet_b0_arch_params
  - checkpoint_params: default_checkpoint_params

arch_params:
  num_classes: 1000

train_dataloader: imagenet_train
val_dataloader: imagenet_val


resume: False
training_hyperparams:
  resume: ${resume}

experiment_name: efficientnet_b0_imagenet


ckpt_root_dir:

multi_gpu: DDP
num_gpus: 4

architecture: efficientnet_b0


# THE FOLLOWING PARAMS ARE DIRECTLY USED BY HYDRA
hydra:
  run:
    # Set the output directory (i.e. where .hydra folder that logs all the input params will be generated)
    dir: ${hydra_output_dir:${ckpt_root_dir}, ${experiment_name}}