Modelpublic

philschmid/layoutlm-funsd-tf

sourceHugging Faceupdated 4y agoView on Hugging Face

0likes21downloads

Model Card

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: 0.2280
—Validation Loss: 0.6532
—Train Overall Precision: 0.7218
—Train Overall Recall: 0.7878
—Train Overall F1: 0.7534
—Train Overall Accuracy: 0.8144
—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', 'learningrate': 3e-05, 'decay': 0.0, 'beta1': 0.9, 'beta2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weightdecay_rate': 0.01}
—trainingprecision: mixedfloat16

Training results

Train Loss	Validation Loss	Train Overall Precision	Train Overall Recall	Train Overall F1	Train Overall Accuracy	Epoch
1.6940	1.4151	0.2686	0.2785	0.2735	0.5128	0
1.1731	0.8665	0.5771	0.6101	0.5932	0.7267	1
0.7612	0.6849	0.6362	0.7336	0.6814	0.7784	2
0.5630	0.6265	0.6748	0.7592	0.7145	0.8017	3
0.4441	0.6256	0.6935	0.7767	0.7328	0.8036	4
0.3641	0.6402	0.7115	0.7772	0.7429	0.7940	5
0.2781	0.6248	0.7176	0.7868	0.7506	0.8141	6
0.2280	0.6532	0.7218	0.7878	0.7534	0.8144	7

Framework versions

—Transformers 4.22.2
—TensorFlow 2.10.0
—Datasets 2.5.2
—Tokenizers 0.12.1