Human_action_classifier
This model is a fine-tuned version of google/vit-base-patch16-224-in21K on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5303
- Accuracy: 0.8496
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.4545 | 0.16 | 100 | 1.3145 | 0.6706 |
| 1.2568 | 0.32 | 200 | 1.0387 | 0.7179 |
| 1.3145 | 0.48 | 300 | 1.0027 | 0.7135 |
| 1.0866 | 0.63 | 400 | 0.8883 | 0.7377 |
| 1.0036 | 0.79 | 500 | 0.8973 | 0.7321 |
| 1.1811 | 0.95 | 600 | 0.8048 | 0.7571 |
| 0.9242 | 1.11 | 700 | 0.9095 | 0.7274 |
| 0.9477 | 1.27 | 800 | 0.8037 | 0.7619 |
| 0.8634 | 1.43 | 900 | 0.7938 | 0.7643 |
| 1.0098 | 1.59 | 1000 | 0.7328 | 0.7766 |
| 0.8176 | 1.75 | 1100 | 0.8065 | 0.7516 |
| 0.8072 | 1.9 | 1200 | 0.7768 | 0.7694 |
| 0.7739 | 2.06 | 1300 | 0.7624 | 0.7726 |
| 0.6851 | 2.22 | 1400 | 0.6687 | 0.7940 |
| 0.7496 | 2.38 | 1500 | 0.6806 | 0.7948 |
| 0.7352 | 2.54 | 1600 | 0.6943 | 0.7897 |
| 0.7311 | 2.7 | 1700 | 0.7353 | 0.7714 |
| 0.7181 | 2.86 | 1800 | 0.6831 | 0.7921 |
| 0.5986 | 3.02 | 1900 | 0.6930 | 0.7897 |
| 0.5716 | 3.17 | 2000 | 0.6685 | 0.8048 |
| 0.5218 | 3.33 | 2100 | 0.7152 | 0.7917 |
| 0.8469 | 3.49 | 2200 | 0.6405 | 0.8020 |
| 0.5783 | 3.65 | 2300 | 0.6728 | 0.7956 |
| 0.7202 | 3.81 | 2400 | 0.6007 | 0.8155 |
| 0.5525 | 3.97 | 2500 | 0.6559 | 0.8056 |
| 0.519 | 4.13 | 2600 | 0.5868 | 0.8222 |
| 0.6171 | 4.29 | 2700 | 0.6157 | 0.8103 |
| 0.5401 | 4.44 | 2800 | 0.6120 | 0.8083 |
| 0.6105 | 4.6 | 2900 | 0.5619 | 0.8325 |
| 0.7497 | 4.76 | 3000 | 0.5859 | 0.8302 |
| 0.4856 | 4.92 | 3100 | 0.5833 | 0.8262 |
| 0.4959 | 5.08 | 3200 | 0.5704 | 0.8329 |
| 0.4413 | 5.24 | 3300 | 0.6217 | 0.8190 |
| 0.4513 | 5.4 | 3400 | 0.5750 | 0.8294 |
| 0.3987 | 5.56 | 3500 | 0.5826 | 0.8341 |
| 0.4395 | 5.71 | 3600 | 0.5754 | 0.8385 |
| 0.4669 | 5.87 | 3700 | 0.5653 | 0.8357 |
| 0.4005 | 6.03 | 3800 | 0.5424 | 0.8377 |
| 0.4457 | 6.19 | 3900 | 0.5620 | 0.8393 |
| 0.3693 | 6.35 | 4000 | 0.5411 | 0.8413 |
| 0.2933 | 6.51 | 4100 | 0.5325 | 0.8484 |
| 0.2603 | 6.67 | 4200 | 0.5360 | 0.8476 |
| 0.3364 | 6.83 | 4300 | 0.5303 | 0.8496 |
| 0.3639 | 6.98 | 4400 | 0.5316 | 0.8492 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 6
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.