videomae-base-finetuned-crema-d8
This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.7377
- Accuracy: 0.7644
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: 8
- eval_batch_size: 8
- 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: 5952
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.3492 | 0.13 | 745 | 1.3353 | 0.5127 |
| 0.9541 | 1.13 | 1490 | 0.9004 | 0.6849 |
| 0.7073 | 2.13 | 2235 | 0.7946 | 0.7236 |
| 0.8417 | 3.13 | 2980 | 0.8516 | 0.6876 |
| 0.3899 | 4.13 | 3725 | 0.7319 | 0.7450 |
| 0.3669 | 5.13 | 4470 | 0.7200 | 0.7490 |
| 0.5429 | 6.13 | 5215 | 0.6304 | 0.7864 |
| 0.2831 | 7.12 | 5952 | 0.6373 | 0.7931 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2