params.yaml

runtime:
  all_reduce_alg: null
  batchnorm_spatial_persistent: false
  dataset_num_private_threads: null
  default_shard_dim: -1
  distribution_strategy: mirrored
  enable_xla: false
  gpu_thread_mode: null
  loss_scale: dynamic
  mixed_precision_dtype: float16
  num_cores_per_replica: 1
  num_gpus: 2
  num_packs: 1
  per_gpu_thread_count: 0
  run_eagerly: false
  task_index: -1
  tpu: null
  worker_hosts: null
task:
  gradient_clip_norm: 0.0
  init_checkpoint: ''
  logging_dir: null
  losses:
    l2_weight_decay: 0.0005
    label_smoothing: 0.0
    one_hot: true
  model:
    add_head_batch_norm: false
    backbone:
      darknet:
        model_id: cspdarknet53
      type: darknet
    dropout_rate: 0.0
    input_size: [256, 256, 3]
    norm_activation:
      activation: mish
      norm_epsilon: 0.001
      norm_momentum: 0.99
      use_sync_bn: false
    num_classes: 1001
  train_data:
    block_length: 1
    cache: false
    cycle_length: 10
    deterministic: null
    drop_remainder: true
    dtype: float16
    enable_tf_data_service: false
    global_batch_size: 16
    input_path: ''
    is_training: true
    sharding: true
    shuffle_buffer_size: 100
    tf_data_service_address: null
    tf_data_service_job_name: null
    tfds_as_supervised: false
    tfds_data_dir: ~/tensorflow_datasets/
    tfds_download: true
    tfds_name: imagenet2012
    tfds_skip_decoding_feature: ''
    tfds_split: train
  validation_data:
    block_length: 1
    cache: false
    cycle_length: 10
    deterministic: null
    drop_remainder: false
    dtype: float16
    enable_tf_data_service: false
    global_batch_size: 16
    input_path: ''
    is_training: true
    sharding: true
    shuffle_buffer_size: 100
    tf_data_service_address: null
    tf_data_service_job_name: null
    tfds_as_supervised: false
    tfds_data_dir: ~/tensorflow_datasets/
    tfds_download: true
    tfds_name: imagenet2012
    tfds_skip_decoding_feature: ''
    tfds_split: validation
trainer:
  allow_tpu_summary: false
  best_checkpoint_eval_metric: ''
  best_checkpoint_export_subdir: ''
  best_checkpoint_metric_comp: higher
  checkpoint_interval: 10000
  continuous_eval_timeout: 3600
  eval_tf_function: true
  max_to_keep: 5
  optimizer_config:
    ema: null
    learning_rate:
      polynomial:
        cycle: false
        decay_steps: 9592000
        end_learning_rate: 1.25e-05
        initial_learning_rate: 0.0125
        name: PolynomialDecay
        power: 4.0
      type: polynomial
    optimizer:
      sgd:
        clipnorm: null
        clipvalue: null
        decay: 0.0
        momentum: 0.9
        name: SGD
        nesterov: false
      type: sgd
    warmup:
      linear:
        name: linear
        warmup_learning_rate: 0
        warmup_steps: 8000
      type: linear
  steps_per_loop: 10000
  summary_interval: 10000
  train_steps: 9600000
  train_tf_function: true
  train_tf_while_loop: true
  validation_interval: 10000
  validation_steps: 3200