<!-- 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. -->
ALBARAN
This model is a fine-tuned version of microsoft/layoutlmv3-large on the ALBARAN dataset. It achieves the following results on the evaluation set:
- Loss: 0.1056
- Precision: 0.9254
- Recall: 0.9254
- F1: 0.9254
- Accuracy: 0.9877
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: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 8.33 | 100 | 0.0782 | 0.9403 | 0.9403 | 0.9403 | 0.9893 |
| No log | 16.67 | 200 | 0.0650 | 0.9403 | 0.9403 | 0.9403 | 0.9908 |
| No log | 25.0 | 300 | 0.0960 | 0.9104 | 0.9104 | 0.9104 | 0.9862 |
| No log | 33.33 | 400 | 0.1052 | 0.9104 | 0.9104 | 0.9104 | 0.9862 |
| 0.0746 | 41.67 | 500 | 0.1007 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
| 0.0746 | 50.0 | 600 | 0.1033 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
| 0.0746 | 58.33 | 700 | 0.1047 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
| 0.0746 | 66.67 | 800 | 0.1049 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
| 0.0746 | 75.0 | 900 | 0.1053 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
| 0.0007 | 83.33 | 1000 | 0.1056 | 0.9254 | 0.9254 | 0.9254 | 0.9877 |
Framework versions
- Transformers 4.25.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.2.2
- Tokenizers 0.13.2
