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: 0.8746976294146106
          - name: Recall
            type: recall
            value: 0.904
          - name: F1
            type: f1
            value: 0.8891074502089993
          - name: Accuracy
            type: accuracy
            value: 0.8368167202572347

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: 0.6541
• Precision: 0.8747
• Recall: 0.904
• F1: 0.8891
• Accuracy: 0.8368

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: 6
• 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	1.0	25	1.2831	0.4033	0.4795	0.4381	0.6092
No log	2.0	50	0.8178	0.7266	0.7935	0.7586	0.7748
No log	3.0	75	0.6843	0.7951	0.8345	0.8143	0.7990
No log	4.0	100	0.6317	0.8024	0.861	0.8307	0.8161
No log	5.0	125	0.5964	0.8260	0.897	0.8600	0.8234
No log	6.0	150	0.6050	0.8204	0.87	0.8445	0.8207
No log	7.0	175	0.6281	0.8203	0.8765	0.8475	0.8168
No log	8.0	200	0.6228	0.8449	0.8985	0.8709	0.8235
No log	9.0	225	0.6213	0.8345	0.88	0.8567	0.8266
No log	10.0	250	0.6173	0.8450	0.897	0.8702	0.8357
No log	11.0	275	0.6476	0.8388	0.8895	0.8634	0.8299
No log	12.0	300	0.6359	0.8584	0.8945	0.8761	0.8382
No log	13.0	325	0.6469	0.8759	0.907	0.8912	0.8395
No log	14.0	350	0.6510	0.8729	0.9035	0.8880	0.8373
No log	15.0	375	0.6555	0.8656	0.902	0.8834	0.8354
No log	16.0	400	0.6541	0.8747	0.904	0.8891	0.8368

Framework versions

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