layoutlmv3-finetuned-invoice-2

This model is a fine-tuned version of microsoft/layoutlmv3-base on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1396
  • Precision: 0.7576
  • Recall: 0.8929
  • F1: 0.8197
  • Accuracy: 0.9742

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 4.35 100 0.4241 0.0 0.0 0.0 0.9135
No log 8.7 200 0.2990 0.2353 0.1429 0.1778 0.9239
No log 13.04 300 0.3107 0.5263 0.3571 0.4255 0.9458
No log 17.39 400 0.1345 0.6970 0.8214 0.7541 0.9742
0.2872 21.74 500 0.1396 0.7576 0.8929 0.8197 0.9742
0.2872 26.09 600 0.1673 0.8519 0.8214 0.8364 0.9690
0.2872 30.43 700 0.1784 0.8519 0.8214 0.8364 0.9690
0.2872 34.78 800 0.1401 0.7742 0.8571 0.8136 0.9729
0.2872 39.13 900 0.1480 0.7273 0.8571 0.7869 0.9716
0.0443 43.48 1000 0.1739 0.6970 0.8214 0.7541 0.9703
0.0443 47.83 1100 0.1786 0.7097 0.7857 0.7458 0.9690
0.0443 52.17 1200 0.1832 0.6970 0.8214 0.7541 0.9690
0.0443 56.52 1300 0.1861 0.6389 0.8214 0.7187 0.9690
0.0443 60.87 1400 0.2155 0.6667 0.7143 0.6897 0.9639
0.0198 65.22 1500 0.2087 0.6667 0.7143 0.6897 0.9652
0.0198 69.57 1600 0.1680 0.6970 0.8214 0.7541 0.9703
0.0198 73.91 1700 0.1664 0.6970 0.8214 0.7541 0.9703
0.0198 78.26 1800 0.1795 0.6970 0.8214 0.7541 0.9703
0.0198 82.61 1900 0.1807 0.6970 0.8214 0.7541 0.9703
0.0151 86.96 2000 0.1825 0.6970 0.8214 0.7541 0.9703

Framework versions

  • Transformers 4.27.4
  • Pytorch 2.0.0+cu117
  • Datasets 2.11.0
  • Tokenizers 0.13.3
Downloads last month
9
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.