<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
Rhenus-Annotations-vgg-json
This model is a fine-tuned version of microsoft/layoutlmv3-base on the dataset dataset. It achieves the following results on the evaluation set:
- Loss: 3.3432
- Precision: 0.9892
- Recall: 0.9634
- F1: 0.9761
- Accuracy: 0.5833
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: 1e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2000
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 2.5 | 100 | 2.4965 | 0.9255 | 0.9110 | 0.9182 | 0.5637 |
| No log | 5.0 | 200 | 2.7204 | 0.9838 | 0.9529 | 0.9681 | 0.5752 |
| No log | 7.5 | 300 | 3.0295 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| No log | 10.0 | 400 | 3.1623 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.4467 | 12.5 | 500 | 3.3432 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.4467 | 15.0 | 600 | 3.4314 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.4467 | 17.5 | 700 | 3.5995 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.4467 | 20.0 | 800 | 3.6942 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.4467 | 22.5 | 900 | 3.7672 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.1072 | 25.0 | 1000 | 3.8307 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.1072 | 27.5 | 1100 | 3.9029 | 0.9786 | 0.9581 | 0.9683 | 0.5822 |
| 0.1072 | 30.0 | 1200 | 3.9604 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.1072 | 32.5 | 1300 | 4.0041 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.1072 | 35.0 | 1400 | 4.0384 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0426 | 37.5 | 1500 | 4.0769 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0426 | 40.0 | 1600 | 4.0993 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0426 | 42.5 | 1700 | 4.1221 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0426 | 45.0 | 1800 | 4.1346 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0426 | 47.5 | 1900 | 4.1458 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
| 0.0205 | 50.0 | 2000 | 4.1482 | 0.9892 | 0.9634 | 0.9761 | 0.5833 |
Framework versions
- Transformers 4.24.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.1
