Edit model card

layoutlmv3-finetuned-invoice

This model is a fine-tuned version of microsoft/layoutlmv3-base on the generated dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0028
  • Precision: 0.9960
  • Recall: 0.9980
  • F1: 0.9970
  • Accuracy: 0.9996

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 2.0 100 0.0502 0.97 0.9838 0.9768 0.9968
No log 4.0 200 0.0194 0.972 0.9858 0.9789 0.9971
No log 6.0 300 0.0160 0.972 0.9858 0.9789 0.9971
No log 8.0 400 0.0123 0.972 0.9858 0.9789 0.9971
0.053 10.0 500 0.0089 0.9757 0.9757 0.9757 0.9966
0.053 12.0 600 0.0058 0.9959 0.9919 0.9939 0.9992
0.053 14.0 700 0.0046 0.9939 0.9919 0.9929 0.9989
0.053 16.0 800 0.0037 0.9960 0.9980 0.9970 0.9996
0.053 18.0 900 0.0068 0.9959 0.9878 0.9919 0.9987
0.0057 20.0 1000 0.0054 0.9919 0.9959 0.9939 0.9992
0.0057 22.0 1100 0.0057 0.9919 0.9959 0.9939 0.9992
0.0057 24.0 1200 0.0049 0.9919 0.9959 0.9939 0.9992
0.0057 26.0 1300 0.0052 0.9919 0.9959 0.9939 0.9992
0.0057 28.0 1400 0.0030 0.9960 0.9980 0.9970 0.9996
0.0022 30.0 1500 0.0028 0.9960 0.9980 0.9970 0.9996
0.0022 32.0 1600 0.0030 0.9960 0.9980 0.9970 0.9996
0.0022 34.0 1700 0.0030 0.9960 0.9980 0.9970 0.9996
0.0022 36.0 1800 0.0037 0.9960 0.9980 0.9970 0.9996
0.0022 38.0 1900 0.0037 0.9960 0.9980 0.9970 0.9996
0.0017 40.0 2000 0.0037 0.9960 0.9980 0.9970 0.9996

Framework versions

  • Transformers 4.23.1
  • Pytorch 1.12.1+cu113
  • Datasets 2.6.1
  • Tokenizers 0.13.1
Downloads last month
2
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.

Evaluation results