metadata

license: cc-by-nc-sa-4.0
base_model: microsoft/layoutlmv3-base
tags:
  - generated_from_trainer
datasets:
  - my_csv_dataset3
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: passive_invoices_v4.7_refined
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: my_csv_dataset3
          type: my_csv_dataset3
          config: discharge
          split: test
          args: discharge
        metrics:
          - name: Precision
            type: precision
            value: 0.8837680590965549
          - name: Recall
            type: recall
            value: 0.9081687491602848
          - name: F1
            type: f1
            value: 0.895802272802571
          - name: Accuracy
            type: accuracy
            value: 0.9791788856304985

passive_invoices_v4.7_refined

This model is a fine-tuned version of microsoft/layoutlmv3-base on the my_csv_dataset3 dataset. It achieves the following results on the evaluation set:

Loss: 0.0915
Precision: 0.8838
Recall: 0.9082
F1: 0.8958
Accuracy: 0.9792

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: 16000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
1.0499	0.27	500	0.8340	0.1650	0.0685	0.0968	0.7864
0.6058	0.53	1000	0.5578	0.3949	0.3288	0.3588	0.8551
0.4061	0.8	1500	0.3891	0.5604	0.5187	0.5388	0.8984
0.2779	1.07	2000	0.3063	0.6178	0.6270	0.6223	0.9156
0.2234	1.33	2500	0.2566	0.6489	0.6511	0.6500	0.9244
0.185	1.6	3000	0.2230	0.7019	0.7136	0.7077	0.9381
0.1524	1.87	3500	0.2003	0.7038	0.7484	0.7254	0.9433
0.1249	2.14	4000	0.1652	0.7548	0.7728	0.7637	0.9546
0.1101	2.4	4500	0.1480	0.7760	0.7986	0.7872	0.9589
0.1054	2.67	5000	0.1455	0.7852	0.8163	0.8004	0.9601
0.0846	2.94	5500	0.1413	0.7828	0.8261	0.8039	0.9610
0.0822	3.2	6000	0.1285	0.8133	0.8213	0.8173	0.9649
0.0725	3.47	6500	0.1256	0.8112	0.8444	0.8275	0.9670
0.0653	3.74	7000	0.1210	0.8178	0.8552	0.8361	0.9673
0.0682	4.0	7500	0.1123	0.8347	0.8624	0.8483	0.9703
0.0562	4.27	8000	0.1084	0.8439	0.8635	0.8536	0.9723
0.0553	4.54	8500	0.1098	0.8323	0.8761	0.8536	0.9710
0.0527	4.81	9000	0.1035	0.8408	0.8819	0.8609	0.9732
0.0446	5.07	9500	0.1037	0.8594	0.8839	0.8715	0.9747
0.047	5.34	10000	0.1080	0.8631	0.8825	0.8727	0.9731
0.0402	5.61	10500	0.0955	0.8696	0.8871	0.8783	0.9768
0.0428	5.87	11000	0.0948	0.8685	0.8957	0.8819	0.9765
0.0422	6.14	11500	0.0992	0.8724	0.8957	0.8839	0.9762
0.0365	6.41	12000	0.0951	0.8731	0.9032	0.8879	0.9777
0.0351	6.67	12500	0.0930	0.8818	0.9018	0.8917	0.9786
0.0353	6.94	13000	0.0973	0.8654	0.9010	0.8828	0.9765
0.0304	7.21	13500	0.0946	0.8795	0.9053	0.8923	0.9784
0.0324	7.47	14000	0.0954	0.8805	0.9048	0.8925	0.9782
0.0327	7.74	14500	0.0920	0.8825	0.9048	0.8935	0.9786
0.0293	8.01	15000	0.0916	0.8810	0.9068	0.8937	0.9789
0.0259	8.28	15500	0.0921	0.8823	0.9062	0.8941	0.9790
0.0337	8.54	16000	0.0915	0.8838	0.9082	0.8958	0.9792

Framework versions

Transformers 4.39.3
Pytorch 2.1.0+cu121
Datasets 2.18.0
Tokenizers 0.15.2