layoutlm-CC-7

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

• Loss: 4.1612
• Precision: 0.1253
• Recall: 0.2022
• F1: 0.1547
• Accuracy: 0.1965

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
4.8141	1.0	1	4.7205	0.0921	0.1311	0.1082	0.0821
4.7028	2.0	2	4.6365	0.1414	0.2022	0.1664	0.1425
4.6011	3.0	3	4.5617	0.1230	0.2022	0.1530	0.1274
4.5126	4.0	4	4.4931	0.1174	0.2022	0.1486	0.1231
4.4376	5.0	5	4.4390	0.1166	0.2022	0.1479	0.1166
4.3778	6.0	6	4.3926	0.1166	0.2022	0.1479	0.1188
4.3224	7.0	7	4.3454	0.1166	0.2022	0.1479	0.1210
4.2658	8.0	8	4.3058	0.1166	0.2022	0.1479	0.1253
4.2182	9.0	9	4.2708	0.1179	0.2022	0.1490	0.1425
4.1796	10.0	10	4.2415	0.1208	0.2022	0.1513	0.1641
4.1423	11.0	11	4.2165	0.1222	0.2022	0.1523	0.1728
4.1197	12.0	12	4.1951	0.1230	0.2022	0.1530	0.1793
4.0976	13.0	13	4.1782	0.1241	0.2022	0.1538	0.1922
4.0801	14.0	14	4.1669	0.1253	0.2022	0.1547	0.1965
4.0627	15.0	15	4.1612	0.1253	0.2022	0.1547	0.1965

Framework versions

• Transformers 4.47.0.dev0
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3