Ammar-alhaj-ali's picture
Update README.md
29b538e
|
raw
history blame
2.2 kB

LayoutLMv3-Fine-Tuning-Invoice Model

Model description

LayoutLMv3-Fine-Tuning-Invoice Model is a fine-tuned version of 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 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 0.070030 0.972000 0.985801 0.978852 0.997051
No log 200 0.017637 0.972000 0.985801 0.978852 0.997051
No log 300 0.015573 0.972000 0.985801 0.978852 0.997051
No log 400 0.011000 0.973737 0.977688 0.978852 0.996419
0.110800 500 0.005334 1.0 1.0 1.0 1.0
0.110800 600 0.002994 1.0 1.0 1.0 1.0
0.110800 700 0.002330 1.0 1.0 1.0 1.0
0.110800 800 0.002188 1.0 1.0 1.0 1.0
0.110800 900 0.002105 1.0 1.0 1.0 1.0
0.004900 1000 0.002111 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