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: 0.2509
• Validation Loss: 0.6942
• Train Overall Precision: 0.7291
• Train Overall Recall: 0.7888
• Train Overall F1: 0.7578
• Train Overall Accuracy: 0.8067
• 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: {'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}
• training_precision: mixed_float16

Training results

Train Loss	Validation Loss	Train Overall Precision	Train Overall Recall	Train Overall F1	Train Overall Accuracy	Epoch
1.6886	1.4100	0.2324	0.2313	0.2318	0.5009	0
1.1702	0.8486	0.5971	0.6618	0.6278	0.7338	1
0.7521	0.7032	0.6561	0.7341	0.6929	0.7687	2
0.5727	0.6268	0.6736	0.7662	0.7169	0.7957	3
0.4586	0.6322	0.6909	0.7772	0.7315	0.7999	4
0.3725	0.6378	0.7134	0.7782	0.7444	0.8096	5
0.2987	0.6835	0.7270	0.7777	0.7515	0.8056	6
0.2509	0.6942	0.7291	0.7888	0.7578	0.8067	7

Framework versions

• Transformers 4.23.1
• TensorFlow 2.6.0
• Datasets 2.6.1
• Tokenizers 0.13.1