imagenet_resnetrs50_i160.yaml

# ResNet-RS-50 ImageNet classification. 79.1% top-1 accuracy.
runtime:
  distribution_strategy: 'tpu'
  mixed_precision_dtype: 'bfloat16'
task:
  model:
    num_classes: 1001
    input_size: [160, 160, 3]
    backbone:
      type: 'resnet'
      resnet:
        model_id: 50
        replace_stem_max_pool: true
        resnetd_shortcut: true
        se_ratio: 0.25
        stem_type: 'v1'
        stochastic_depth_drop_rate: 0.0
    norm_activation:
      activation: 'swish'
      norm_momentum: 0.0
      use_sync_bn: false
    dropout_rate: 0.25
  losses:
    l2_weight_decay: 0.00004
    one_hot: true
    label_smoothing: 0.1
  train_data:
    input_path: 'gs://mlcompass-data/imagenet/imagenet-2012-tfrecord/train*'
    is_training: true
    global_batch_size: 4096
    dtype: 'bfloat16'
    aug_type:
      type: 'randaug'
      randaug:
        magnitude: 10
  validation_data:
    input_path: 'gs://mlcompass-data/imagenet/imagenet-2012-tfrecord/valid*'
    is_training: false
    global_batch_size: 4096
    dtype: 'bfloat16'
    drop_remainder: false
trainer:
  train_steps: 109200
  validation_steps: 13
  validation_interval: 312
  steps_per_loop: 312
  summary_interval: 312
  checkpoint_interval: 312
  optimizer_config:
    ema:
      average_decay: 0.9999
    optimizer:
      type: 'sgd'
      sgd:
        momentum: 0.9
    learning_rate:
      type: 'cosine'
      cosine:
        initial_learning_rate: 1.6
        decay_steps: 109200
    warmup:
      type: 'linear'
      linear:
        warmup_steps: 1560