README.md
| 1 | --- |
| 2 | library_name: transformers |
| 3 | license: cc-by-nc-4.0 |
| 4 | base_model: MCG-NJU/videomae-small-finetuned-kinetics |
| 5 | tags: |
| 6 | - generated_from_trainer |
| 7 | metrics: |
| 8 | - accuracy |
| 9 | model-index: |
| 10 | - name: videomae-small-finetuned-kinetics-finetuned-xd-violence |
| 11 | results: [] |
| 12 | --- |
| 13 | |
| 14 | <!-- This model card has been generated automatically according to the information the Trainer had access to. You |
| 15 | should probably proofread and complete it, then remove this comment. --> |
| 16 | |
| 17 | # videomae-small-finetuned-kinetics-finetuned-xd-violence |
| 18 | |
| 19 | This model is a fine-tuned version of [MCG-NJU/videomae-small-finetuned-kinetics](https://huggingface.co/MCG-NJU/videomae-small-finetuned-kinetics) on an unknown dataset. |
| 20 | It achieves the following results on the evaluation set: |
| 21 | - Loss: 0.3828 |
| 22 | - Accuracy: 0.8254 |
| 23 | |
| 24 | |
| 25 | Video preprocessing config: |
| 26 | Resize to: (224, 224) |
| 27 | Frames to sample: 16 |
| 28 | Clip duration: 2.1333333333333333s |
| 29 | |
| 30 | ## Model description |
| 31 | |
| 32 | More information needed |
| 33 | |
| 34 | ## Intended uses & limitations |
| 35 | |
| 36 | More information needed |
| 37 | |
| 38 | ## Training and evaluation data |
| 39 | |
| 40 | More information needed |
| 41 | |
| 42 | ## Training procedure |
| 43 | |
| 44 | ### Training hyperparameters |
| 45 | |
| 46 | The following hyperparameters were used during training: |
| 47 | - learning_rate: 5e-05 |
| 48 | - train_batch_size: 2 |
| 49 | - eval_batch_size: 2 |
| 50 | - seed: 42 |
| 51 | - gradient_accumulation_steps: 4 |
| 52 | - total_train_batch_size: 8 |
| 53 | - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
| 54 | - lr_scheduler_type: linear |
| 55 | - lr_scheduler_warmup_ratio: 0.1 |
| 56 | - num_epochs: 3 |
| 57 | |
| 58 | ### Training results |
| 59 | |
| 60 | | Training Loss | Epoch | Step | Validation Loss | Accuracy | |
| 61 | |:-------------:|:-----:|:----:|:---------------:|:--------:| |
| 62 | | 0.3574 | 1.0 | 394 | 0.5181 | 0.7738 | |
| 63 | | 0.4428 | 2.0 | 788 | 0.4357 | 0.8119 | |
| 64 | | 0.075 | 3.0 | 1182 | 0.4100 | 0.8183 | |
| 65 | |
| 66 | |
| 67 | ### Testing results |
| 68 | |
| 69 | ``` |
| 70 | {'eval_loss': 0.3827999532222748, |
| 71 | 'eval_accuracy': 0.8253768844221105, |
| 72 | 'eval_runtime': 184.3164, |
| 73 | 'eval_samples_per_second': 4.319, |
| 74 | 'eval_steps_per_second': 2.159, |
| 75 | 'epoch': 3.0} |
| 76 | ``` |
| 77 | |
| 78 | | Class | Precision | Recall | F1-Score | Support | |
| 79 | | :--- | :--- | :--- | :--- | :--- | |
| 80 | | **Normal (0)** | 0.74 | 0.83 | 0.78 | 299 | |
| 81 | | **Violent (1)** | 0.89 | 0.82 | 0.85 | 497 | |
| 82 | | | | | | | |
| 83 | | **Accuracy** | | | 0.83 | 796 | |
| 84 | | **Macro Avg** | 0.81 | 0.83 | 0.82 | 796 | |
| 85 | | **Weighted Avg** | 0.83 | 0.83 | 0.83 | 796 | |
| 86 | |
| 87 | ### Framework versions |
| 88 | |
| 89 | - Transformers 4.57.1 |
| 90 | - Pytorch 2.8.0+cu126 |
| 91 | - Datasets 4.0.0 |
| 92 | - Tokenizers 0.22.1 |
| 93 | |