| ## LayoutLMv3-Fine-Tuning-Invoice Model | |
| #### Model description | |
| **LayoutLMv3-Fine-Tuning-Invoice Model** is a fine-tuned version of [microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) on the invoice dataset. For the fine-tuning, We used [Invoice Dataset] that includes 12 labels ('Other', 'ABN', 'BILLER', 'BILLER_ADDRESS', 'BILLER_POST_CODE', 'DUE_DATE', 'GST', 'INVOICE_DATE', 'INVOICE_NUMBER', 'SUBTOTAL', 'TOTAL', 'BILLER_ADDRESS'). | |
| This model is a fine-tuned version of [microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) on the invoice dataset. | |
| It achieves the following results on the evaluation set: | |
| - Loss: 0.005334 | |
| - Precision: 1.0 | |
| - Recall: 1.0 | |
| - F1: 1.0 | |
| - Accuracy: 1.0 | |
| ## Training procedure | |
| ### Training hyperparameters | |
| The following hyperparameters were used during training: | |
| - learning_rate: 1.5e-05 | |
| - train_batch_size: 2 | |
| - eval_batch_size: 2 | |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 | |
| - lr_scheduler_type: cosine | |
| - training_steps: 1000 | |
| ### Training results | |
| | Training Loss | Step | Validation Loss | Precision | Recall | F1 | Accuracy | | |
| |:-------------:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| | |
| | No log | 100 | 0.0878 | 0.968 | 0.9817 | 0.9748 | 0.9966 | | |
| | No log | 200 | 0.0241 | 0.972 | 0.9858 | 0.9789 | 0.9971 | | |
| | No log | 300 | 0.0186 | 0.972 | 0.9858 | 0.9789 | 0.9971 | | |
| | No log | 400 | 0.0184 | 0.9854 | 0.9574 | 0.9712 | 0.9956 | | |
| | 0.110800 | 500 | 0.0016 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| | 0.110800 | 600 | 0.0015 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| | 0.110800 | 700 | 0.0014 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| | 0.110800 | 800 | 0.0013 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| | 0.110800 | 900 | 0.0012 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| | 0.004900 | 1000 | 0.0012 | 1.0 | 1.0 | 1.0 | 1.0 | | |
| ### Framework versions | |
| - Transformers 4.20.0.dev0 | |
| - Pytorch 1.11.0+cu113 | |
| - Datasets 2.2.2 | |
| - Tokenizers 0.12.1 |