data: loader: prefetch_size: 4 shuffle_file: False shuffle_buffer: 10000 cycle_length: 32 block_length: 16 max_instances_per_image: 100 skip_crowd_during_training: True image_size: '512x512' num_classes: 91 train_tfrecords: - '/home/ubuntu/workdir/nvidia/tao/tao-experiments/data/train-*' val_tfrecords: - '/home/ubuntu/workdir/nvidia/tao/tao-experiments/data/val-*' val_json_file: '/home/ubuntu/workdir/nvidia/tao/tao-experiments/data/raw-data/annotations/instances_val2017.json' train: optimizer: name: 'sgd' momentum: 0.9 lr_schedule: name: 'cosine' warmup_epoch: 5 warmup_init: 0.0001 learning_rate: 0.2 amp: False checkpoint: "/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/pretrained_efficientdet_tf2_vefficientnet_b0" num_examples_per_epoch: 16400 moving_average_decay: 0.999 batch_size: 20 checkpoint_interval: 5 l2_weight_decay: 0.00004 l1_weight_decay: 0.0 clip_gradients_norm: 10.0 image_preview: True qat: False random_seed: 42 pruned_model_path: '' num_epochs: 2001 model: name: 'efficientdet-d0' aspect_ratios: '[(1.0, 1.0), (1.4, 0.7), (0.7, 1.4)]' anchor_scale: 4 min_level: 3 max_level: 7 num_scales: 3 freeze_bn: True freeze_blocks: [0, 1, 2, 3, 4, 5, 6, 7] #freeze_blocks: 0 # When using pretrained-weights not all layers need to be retrained #freeze_blocks: 1 #freeze_blocks: 2 #freeze_blocks: 3 #freeze_blocks: 4 #freeze_blocks: 5 #freeze_blocks: 6 augment: rand_hflip: True random_crop_min_scale: 0.1 random_crop_max_scale: 2 evaluate: batch_size: 8 num_samples: 500 max_detections_per_image: 100 label_map: "/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/specs/coco_labels.yaml" model_path: "/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/experiment_dir_unpruned/weights/efficientdet-d0_185.tlt" prune: model_path: "/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/experiment_dir_unpruned/weights/efficientdet-d0_185.tlt" normalizer: 'max' output_path: "/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/experiment_dir_pruned/model_pruned.tlt" equalization_criterion: 'union' granularity: 8 threshold: 0.25 min_num_filters: 16 key: 'nvidia_tlt' results_dir: '/home/ubuntu/workdir/nvidia/tao/tao-experiments/efficientdet_tf2/experiment_dir_unpruned