layoutlmv3_cord_model_trained_on_layoutlmv2_cord_ds

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

• Loss: 0.1581
• Precision: 0.9602
• Recall: 0.9555
• F1: 0.9578
• Accuracy: 0.9634

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-05
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	400	0.5418	0.8409	0.8293	0.8350	0.8550
1.1231	2.0	800	0.2616	0.9262	0.9239	0.9251	0.9405
0.2749	3.0	1200	0.3170	0.9272	0.9280	0.9276	0.9259
0.1533	4.0	1600	0.1518	0.9529	0.9498	0.9514	0.9602
0.0808	5.0	2000	0.1581	0.9602	0.9555	0.9578	0.9634

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1