lilt-en-funsd-org

This model is a fine-tuned version of SCUT-DLVCLab/lilt-roberta-en-base on the funsd-layoutlmv3 dataset. It achieves the following results on the evaluation set:

• Loss: 1.8428
• Answer: {'precision': 0.047225501770956316, 'recall': 0.09791921664626684, 'f1': 0.06371963361210674, 'number': 817}
• Header: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}
• Question: {'precision': 0.08554412560909583, 'recall': 0.2934076137418756, 'f1': 0.13246698805281912, 'number': 1077}
• Overall Precision: 0.0730
• Overall Recall: 0.1967
• Overall F1: 0.1065
• Overall Accuracy: 0.2652

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 3
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2