layoutlmv3-finetuned-cord_100

This model is a fine-tuned version of microsoft/layoutlmv3-base on the layoutlm_v3 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2976
• Precision: 0.9298
• Recall: 0.9416
• F1: 0.9357
• Accuracy: 0.9393

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: 1e-05
• train_batch_size: 5
• eval_batch_size: 5
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 2500

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	4.17	250	1.0222	0.7468	0.7949	0.7701	0.8014
1.3962	8.33	500	0.5292	0.8414	0.8735	0.8571	0.8778
1.3962	12.5	750	0.3844	0.9049	0.9192	0.9120	0.9249
0.335	16.67	1000	0.3302	0.9243	0.9326	0.9285	0.9342
0.335	20.83	1250	0.3062	0.9204	0.9349	0.9276	0.9406
0.1419	25.0	1500	0.2931	0.9268	0.9386	0.9327	0.9414
0.1419	29.17	1750	0.2925	0.9248	0.9386	0.9316	0.9359
0.0801	33.33	2000	0.2963	0.9276	0.9394	0.9334	0.9359
0.0801	37.5	2250	0.2916	0.9283	0.9401	0.9342	0.9363
0.0584	41.67	2500	0.2976	0.9298	0.9416	0.9357	0.9393

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2