layoutlmv3-finetuned-letter_100

This model is a fine-tuned version of microsoft/layoutlmv3-base on the data_cartas_layoutv3 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1884
• Precision: 0.7412
• Recall: 0.8673
• F1: 0.7993
• Accuracy: 0.9632

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-06
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 3000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	3.57	250	0.4503	0.2934	0.2242	0.2542	0.8928
0.5521	7.14	500	0.2833	0.4291	0.4639	0.4458	0.9209
0.5521	10.71	750	0.2116	0.5702	0.6753	0.6183	0.9437
0.173	14.29	1000	0.1786	0.6414	0.7835	0.7053	0.9562
0.173	17.86	1250	0.1772	0.6815	0.8492	0.7562	0.9581
0.077	21.43	1500	0.1737	0.7144	0.8737	0.7861	0.9616
0.077	25.0	1750	0.1768	0.7311	0.8724	0.7955	0.9615
0.0441	28.57	2000	0.1694	0.7726	0.8273	0.7990	0.9646
0.0441	32.14	2250	0.1874	0.7400	0.8621	0.7964	0.9620
0.0293	35.71	2500	0.1862	0.7321	0.8698	0.7951	0.9622
0.0293	39.29	2750	0.1887	0.7332	0.8711	0.7962	0.9620
0.0237	42.86	3000	0.1884	0.7412	0.8673	0.7993	0.9632

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3