layoutlm-funsd

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

• Loss: 1.1493
• Answer: {'precision': 0.22598870056497175, 'recall': 0.19777503090234858, 'f1': 0.2109426499670402, 'number': 809}
• Header: {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}
• Question: {'precision': 0.5350523771152297, 'recall': 0.6234741784037559, 'f1': 0.5758889852558543, 'number': 1065}
• Overall Precision: 0.4228
• Overall Recall: 0.4134
• Overall F1: 0.4181
• Overall Accuracy: 0.6373

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Answer	Header	Question	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
1.6256	1.0	10	1.4524	{'precision': 0.05670665212649945, 'recall': 0.06427688504326329, 'f1': 0.060254924681344156, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.3111913357400722, 'recall': 0.40469483568075115, 'f1': 0.35183673469387755, 'number': 1065}	0.2098	0.2423	0.2249	0.4826
1.3478	2.0	20	1.2324	{'precision': 0.14285714285714285, 'recall': 0.1211372064276885, 'f1': 0.1311036789297659, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.4789156626506024, 'recall': 0.5971830985915493, 'f1': 0.5315503552026745, 'number': 1065}	0.3644	0.3683	0.3664	0.5977
1.155	3.0	30	1.1493	{'precision': 0.22598870056497175, 'recall': 0.19777503090234858, 'f1': 0.2109426499670402, 'number': 809}	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 119}	{'precision': 0.5350523771152297, 'recall': 0.6234741784037559, 'f1': 0.5758889852558543, 'number': 1065}	0.4228	0.4134	0.4181	0.6373

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cpu
• Datasets 2.14.4
• Tokenizers 0.13.3