LayoutLMv3_5_entities_4

This model is a fine-tuned version of microsoft/layoutlmv3-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2514
• Precision: 0.8762
• Recall: 0.8519
• F1: 0.8638
• Accuracy: 0.9739

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-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 2000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	2.56	100	0.2241	0.8571	0.8333	0.8451	0.9691
No log	5.13	200	0.2210	0.8952	0.8704	0.8826	0.9758
No log	7.69	300	0.2300	0.9029	0.8611	0.8815	0.9758
No log	10.26	400	0.2630	0.8922	0.8426	0.8667	0.9720
0.0021	12.82	500	0.2692	0.8980	0.8148	0.8544	0.9710
0.0021	15.38	600	0.2414	0.9	0.8333	0.8654	0.9729
0.0021	17.95	700	0.2617	0.875	0.8426	0.8585	0.9729
0.0021	20.51	800	0.2558	0.8713	0.8148	0.8421	0.9720
0.0021	23.08	900	0.2581	0.8725	0.8241	0.8476	0.9729
0.0006	25.64	1000	0.2574	0.8679	0.8519	0.8598	0.9739
0.0006	28.21	1100	0.2806	0.88	0.8148	0.8462	0.9710
0.0006	30.77	1200	0.3032	0.8958	0.7963	0.8431	0.9691
0.0006	33.33	1300	0.2627	0.8889	0.8148	0.8502	0.9729
0.0006	35.9	1400	0.2661	0.8980	0.8148	0.8544	0.9720
0.0006	38.46	1500	0.2650	0.9	0.8333	0.8654	0.9739
0.0006	41.03	1600	0.2543	0.8835	0.8426	0.8626	0.9729
0.0006	43.59	1700	0.2593	0.8911	0.8333	0.8612	0.9739
0.0006	46.15	1800	0.2494	0.8857	0.8611	0.8732	0.9749
0.0006	48.72	1900	0.2494	0.8857	0.8611	0.8732	0.9749
0.0002	51.28	2000	0.2514	0.8762	0.8519	0.8638	0.9739

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3