How to add weight bias in TAO config file?

I am using the classification_tf2 tao_voc with this config:

results_dir: '/workspace/tao-experiments/classification_tf2/output'
dataset:
  train_dataset_path: "/workspace/tao-experiments/data/train"
  val_dataset_path: "/workspace/tao-experiments/data/val"
  preprocess_mode: 'torch'
  num_classes: 2
  augmentation:
    enable_color_augmentation: True
    enable_center_crop: True

train:
  #class_weights: [1.0, 2.0]  # Moderate bias toward class_2
  qat: False
  checkpoint: ''
  batch_size_per_gpu: 32
  num_epochs: 100
  optim_config:
    optimizer: 'sgd'
  lr_config:
    scheduler: 'cosine'
    learning_rate: 0.05
    soft_start: 0.05
  reg_config:
    type: 'L2'
    scope: ['conv2d', 'dense']
    weight_decay: 0.00005
model:
  backbone: 'efficientnet-b0'
  input_width: 256
  input_height: 256
  input_channels: 3
  input_image_depth: 8
evaluate:
  dataset_path: "/workspace/tao-experiments/data/test"
  checkpoint: "/workspace/tao-experiments/classification_tf2/output/train/efficientnet-b0_100.tlt"
  top_k: 3
  batch_size: 256
  n_workers: 8
prune:
  checkpoint: '/workspace/tao-experiments/classification_tf2/output/train/efficientnet-b0_100.tlt'
  threshold: 0.68
  byom_model_path: ''

How do I add weight_bias here?

4 posts - 2 participants

Read full topic