README.md

metadata

tags:
  - generated_from_trainer
datasets:
  - nielsr/funsd-layoutlmv3
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: layoutlmv3-finetuned-funsd
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: nielsr/funsd-layoutlmv3
          type: nielsr/funsd-layoutlmv3
          args: funsd
        metrics:
          - name: Precision
            type: precision
            value: 0.918177
          - name: Recall
            type: recall
            value: 0.930949
          - name: F1
            type: f1
            value: 0.924519
          - name: Accuracy
            type: accuracy
            value: 0.859979

layoutlmv3-finetuned-funsd

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

• Loss: 1.038318
• Precision: 0.918177
• Recall: 0.930949
• F1: 0.924519
• Accuracy: 0.859979