nodes: 8
tasks_per_node: 8
cpus_per_task: 12
mem_per_gpu: 200G
tag: diving48-vitl16-32x4x3
eval_name: video_classification_frozen
folder: /your_folder/evals/vitl/diving48
resume_checkpoint: true
experiment:
  classifier:
    num_probe_blocks: 4
    num_heads: 16
  data:
    dataset_type: VideoDataset
    dataset_train: /your_data_dir/diving48/annotations/Diving48_train_paths.csv
    dataset_val: /your_data_dir/diving48/annotations/Diving48_test_paths.csv
    num_classes: 48
    resolution: 256
    frames_per_clip: 32
    frame_step: 2
    num_segments: 4
    num_views_per_segment: 3
  optimization:
    use_pos_embed: false
    num_epochs: 100
    batch_size: 2
    use_bfloat16: true
    multihead_kwargs:
      - weight_decay: 0.8
        final_weight_decay: 0.8
        lr: 0.001
        start_lr: 0.001
        final_lr: 0.0
        warmup: 0.
      - weight_decay: 0.8
        final_weight_decay: 0.8
        lr: 0.0003
        start_lr: 0.0003
        final_lr: 0.0
        warmup: 0.
      - weight_decay: 0.8
        final_weight_decay: 0.8
        lr: 0.0001
        start_lr: 0.0001
        final_lr: 0.0
        warmup: 0.
model_kwargs:
  checkpoint: /your_vjepa2_checkpoints/vitl.pt
  module_name: evals.video_classification_frozen.modelcustom.vit_encoder_multiclip_multilevel
  wrapper_kwargs:
    max_frames: 128
    use_pos_embed: false
    out_layers: [17, 19, 21, 23]
  pretrain_kwargs:
    encoder:
      model_name: vit_large
      checkpoint_key: target_encoder
      tubelet_size: 2
      patch_size: 16
      uniform_power: true
      use_rope: true