LayoutLMv3_97_2

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: 0.5892
• Precision: 0.8315
• Recall: 0.7721
• F1: 0.8007
• Accuracy: 0.9122

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: 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	0.4807	0.6058	0.4966	0.5458	0.8297
No log	5.13	200	0.3940	0.7553	0.6088	0.6742	0.8771
No log	7.69	300	0.3804	0.7438	0.7109	0.7270	0.9008
No log	10.26	400	0.3900	0.8185	0.8129	0.8157	0.9096
0.2035	12.82	500	0.4102	0.8255	0.7721	0.7979	0.9087
0.2035	15.38	600	0.4077	0.8095	0.8095	0.8095	0.9148
0.2035	17.95	700	0.4915	0.7867	0.7653	0.7759	0.8982
0.2035	20.51	800	0.4861	0.8269	0.7959	0.8111	0.9131
0.2035	23.08	900	0.5051	0.7818	0.7313	0.7557	0.9052
0.0117	25.64	1000	0.5404	0.8303	0.7653	0.7965	0.9069
0.0117	28.21	1100	0.6110	0.8492	0.7279	0.7839	0.9061
0.0117	30.77	1200	0.5379	0.8014	0.7823	0.7917	0.9096
0.0117	33.33	1300	0.5343	0.8057	0.7755	0.7903	0.9131
0.0117	35.9	1400	0.5590	0.8333	0.7653	0.7979	0.9140
0.0013	38.46	1500	0.6296	0.8488	0.7449	0.7935	0.9122
0.0013	41.03	1600	0.6089	0.8421	0.7619	0.8	0.9122
0.0013	43.59	1700	0.5869	0.8291	0.7755	0.8014	0.9140
0.0013	46.15	1800	0.5847	0.8291	0.7755	0.8014	0.9140
0.0013	48.72	1900	0.5881	0.8285	0.7721	0.7993	0.9131
0.0004	51.28	2000	0.5892	0.8315	0.7721	0.8007	0.9122

Framework versions

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