videomae-base-finetuned-kinetics-finetuned-fall-detect
This model is a fine-tuned version of MCG-NJU/videomae-base-finetuned-kinetics on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1793
- Accuracy: 0.9277
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 1452
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.6416 | 0.12 | 180 | 0.4387 | 0.8874 |
| 0.4418 | 1.12 | 360 | 0.3639 | 0.8760 |
| 0.4265 | 2.12 | 540 | 0.2258 | 0.9201 |
| 0.3196 | 3.12 | 720 | 0.2528 | 0.9201 |
| 0.3722 | 4.12 | 900 | 0.1813 | 0.9413 |
| 0.2537 | 5.12 | 1080 | 0.1401 | 0.9625 |
| 0.0953 | 6.12 | 1260 | 0.1458 | 0.9527 |
| 0.2226 | 7.12 | 1440 | 0.1575 | 0.9494 |
| 0.1424 | 8.01 | 1452 | 0.1595 | 0.9511 |
Framework versions
- Transformers 4.39.0
- Pytorch 2.2.1+cu118
- Datasets 2.18.0
- Tokenizers 0.15.2