layoutlm-funsd-tf

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

• Train Loss: 0.5754
• Validation Loss: 1.0073
• Train Overall Precision: 0.4858
• Train Overall Recall: 0.5735
• Train Overall F1: 0.5260
• Train Overall Accuracy: 0.6411
• Epoch: 7

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:

• optimizer: {'inner_optimizer': {'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': 3e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
• training_precision: mixed_float16

Training results

Train Loss	Validation Loss	Train Overall Precision	Train Overall Recall	Train Overall F1	Train Overall Accuracy	Epoch
1.6595	1.4718	0.1431	0.2780	0.1889	0.4016	0
1.3342	1.2762	0.2962	0.4942	0.3704	0.4644	1
1.1464	1.1828	0.3753	0.5173	0.4350	0.5034	2
1.0198	1.0195	0.4070	0.5359	0.4626	0.6167	3
0.8729	1.0543	0.4343	0.5740	0.4945	0.6018	4
0.7979	1.2603	0.4648	0.5866	0.5186	0.5615	5
0.6799	1.0257	0.5180	0.5775	0.5461	0.6408	6
0.5754	1.0073	0.4858	0.5735	0.5260	0.6411	7

Framework versions

• Transformers 4.33.3
• TensorFlow 2.10.0
• Datasets 2.16.1
• Tokenizers 0.13.2