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

# Checkout the datasets at https://universe.roboflow.com/roboflow-100?ref=blog.roboflow.com
#
# `dataset_name` refers to the official name of the dataset.
# You can find it in the url of the dataset: https://universe.roboflow.com/roboflow-100/digits-t2eg6 -> digits-t2eg6
#
# Example: python train_from_recipe.py --config-name=roboflow_yolox dataset_name=digits-t2eg6

defaults:
  - training_hyperparams: coco2017_yolox_train_params
  - dataset_params: roboflow_detection_dataset_params
  - checkpoint_params: default_checkpoint_params
  - _self_

dataset_name: ??? # Placeholder for the name of the dataset you want to use (e.g. "digits-t2eg6")
dataset_params:
  dataset_name: ${dataset_name}
num_classes: ${roboflow_dataset_num_classes:${dataset_name}}


architecture: yolox_m
arch_params:
  num_classes: ${num_classes}
  yolo_type: 'yoloX'
  depth_mult_factor: 0.67
  width_mult_factor: 0.75


train_dataloader: roboflow_train_yolox
val_dataloader: roboflow_val_yolox


load_checkpoint: False
checkpoint_params:
  pretrained_weights: coco

result_path: # By defaults saves results in checkpoints directory
resume: False
training_hyperparams:
  max_epochs: 100
  resume: ${resume}
  criterion_params:
    num_classes: ${num_classes}
  train_metrics_list:
    - DetectionMetrics:
        normalize_targets: True
        post_prediction_callback:
          _target_: super_gradients.training.models.detection_models.yolo_base.YoloPostPredictionCallback
          iou: 0.65
          conf: 0.01
        num_cls: 80
  valid_metrics_list:
    - DetectionMetrics:
        normalize_targets: True
        post_prediction_callback:
          _target_: super_gradients.training.models.detection_models.yolo_base.YoloPostPredictionCallback
          iou: 0.65
          conf: 0.01
        num_cls: 80


multi_gpu: DDP
num_gpus: 3
experiment_name: ${architecture}_roboflow_${dataset_name}
ckpt_root_dir:


# 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}}