README.md

metadata

tags:
  - generated_from_trainer
datasets:
  - mp-02/funsd
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: layoutlmv3-finetuned-funsd
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: mp-02/funsd
          type: mp-02/funsd
        metrics:
          - name: Precision
            type: precision
            value: 0.8553875236294896
          - name: Recall
            type: recall
            value: 0.905
          - name: F1
            type: f1
            value: 0.8794946550048591
          - name: Accuracy
            type: accuracy
            value: 0.833371612310519

layoutlmv3-finetuned-funsd

This model is a fine-tuned version of layoutlmv3 on the mp-02/funsd dataset. It achieves the following results on the evaluation set:

• Loss: 0.5784
• Precision: 0.8554
• Recall: 0.905
• F1: 0.8795
• Accuracy: 0.8334

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: 4
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 400

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.66	25	1.3511	0.3301	0.3585	0.3437	0.5721
No log	1.32	50	0.9059	0.6965	0.7515	0.7229	0.7615
No log	1.97	75	0.7164	0.7613	0.831	0.7946	0.7796
No log	2.63	100	0.6393	0.7947	0.8575	0.8249	0.7993
No log	3.29	125	0.5756	0.8138	0.87	0.8410	0.8104
No log	3.95	150	0.5508	0.8197	0.884	0.8506	0.8323
No log	4.61	175	0.5458	0.8325	0.8895	0.8600	0.8328
No log	5.26	200	0.5740	0.8234	0.8765	0.8491	0.8266
No log	5.92	225	0.5719	0.8532	0.8895	0.8710	0.8361
No log	6.58	250	0.5436	0.8439	0.9055	0.8736	0.8264
No log	7.24	275	0.5714	0.8520	0.9065	0.8784	0.8290
No log	7.89	300	0.5853	0.8560	0.9035	0.8791	0.8281
No log	8.55	325	0.5702	0.8578	0.905	0.8808	0.8390
No log	9.21	350	0.5667	0.8552	0.901	0.8775	0.8419
No log	9.87	375	0.5793	0.8552	0.9035	0.8787	0.8338
No log	10.53	400	0.5784	0.8554	0.905	0.8795	0.8334

Framework versions

• Transformers 4.12.5
• Pytorch 1.10.0+cu111
• Datasets 2.13.2
• Tokenizers 0.10.1