| nodes: 8 | |
| tasks_per_node: 8 | |
| tag: k400-16x8x3 | |
| eval_name: video_classification_frozen | |
| resume_checkpoint: false | |
| data: | |
| dataset_train: /your_path_to_kinetics400_train_csv_file_index.csv | |
| dataset_val: /your_path_to_kinetics400_val_csv_file_index.csv | |
| dataset_type: VideoDataset | |
| num_classes: 400 | |
| frames_per_clip: 16 | |
| num_segments: 8 | |
| num_views_per_segment: 3 | |
| frame_step: 4 | |
| optimization: | |
| attend_across_segments: true | |
| num_epochs: 20 | |
| resolution: 224 | |
| batch_size: 4 | |
| weight_decay: 0.01 | |
| lr: 0.001 | |
| start_lr: 0.001 | |
| final_lr: 0.0 | |
| warmup: 0. | |
| use_bfloat16: true | |
| pretrain: | |
| model_name: vit_large | |
| checkpoint_key: target_encoder | |
| clip_duration: null | |
| frames_per_clip: 16 | |
| tubelet_size: 2 | |
| uniform_power: true | |
| use_silu: false | |
| tight_silu: false | |
| use_sdpa: true | |
| patch_size: 16 | |
| folder: /your_absolute_file_path_to_directory_where_pretrained_models_are_contained/ | |
| checkpoint: jepa-latest.pth.tar # name of pretrained model file inside folder | |
| write_tag: jepa | |