<!-- 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. -->
SOCIEDADES
This model is a fine-tuned version of microsoft/layoutlmv3-base on the sroie dataset. It achieves the following results on the evaluation set:
- Loss: 0.0008
- Precision: 0.9723
- Recall: 0.9723
- F1: 0.9723
- Accuracy: 0.9998
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 | 0.28 | 100 | 0.0013 | 0.9686 | 0.9619 | 0.9653 | 0.9998 |
No log | 0.56 | 200 | 0.0025 | 0.9070 | 0.9446 | 0.9254 | 0.9994 |
No log | 0.85 | 300 | 0.0067 | 0.8562 | 0.9273 | 0.8904 | 0.9985 |
No log | 1.13 | 400 | 0.0013 | 0.9454 | 0.9585 | 0.9519 | 0.9997 |
0.0055 | 1.41 | 500 | 0.0009 | 0.9654 | 0.9654 | 0.9654 | 0.9998 |
0.0055 | 1.69 | 600 | 0.0042 | 0.8795 | 0.9343 | 0.9060 | 0.9987 |
0.0055 | 1.97 | 700 | 0.0006 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.0055 | 2.25 | 800 | 0.0010 | 0.9485 | 0.9550 | 0.9517 | 0.9998 |
0.0055 | 2.54 | 900 | 0.0006 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.001 | 2.82 | 1000 | 0.0014 | 0.9553 | 0.9619 | 0.9586 | 0.9997 |
0.001 | 3.1 | 1100 | 0.0010 | 0.9585 | 0.9585 | 0.9585 | 0.9998 |
0.001 | 3.38 | 1200 | 0.0009 | 0.9585 | 0.9585 | 0.9585 | 0.9998 |
0.001 | 3.66 | 1300 | 0.0010 | 0.9585 | 0.9585 | 0.9585 | 0.9998 |
0.001 | 3.94 | 1400 | 0.0010 | 0.9654 | 0.9654 | 0.9654 | 0.9998 |
0.0003 | 4.23 | 1500 | 0.0008 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.0003 | 4.51 | 1600 | 0.0009 | 0.9654 | 0.9654 | 0.9654 | 0.9998 |
0.0003 | 4.79 | 1700 | 0.0008 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.0003 | 5.07 | 1800 | 0.0008 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.0003 | 5.35 | 1900 | 0.0008 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
0.0002 | 5.63 | 2000 | 0.0008 | 0.9723 | 0.9723 | 0.9723 | 0.9998 |
Framework versions
- Transformers 4.27.0.dev0
- Pytorch 1.13.1+cu116
- Datasets 2.2.2
- Tokenizers 0.13.2