Output_LayoutLMv3_v3

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

• Loss: 0.1344
• Precision: 0.7699
• Recall: 0.8142
• F1: 0.7914
• Accuracy: 0.9695

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-07
• train_batch_size: 4
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 3000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	4.55	100	0.5786	0.0	0.0	0.0	0.8867
No log	9.09	200	0.4032	0.0	0.0	0.0	0.8867
No log	13.64	300	0.2908	0.4091	0.1593	0.2293	0.9067
No log	18.18	400	0.2300	0.5858	0.4381	0.5013	0.9267
0.5251	22.73	500	0.1981	0.685	0.6062	0.6432	0.9438
0.5251	27.27	600	0.1790	0.7130	0.6814	0.6968	0.9505
0.5251	31.82	700	0.1689	0.7249	0.7345	0.7297	0.9581
0.5251	36.36	800	0.1593	0.7478	0.7478	0.7478	0.9619
0.5251	40.91	900	0.1582	0.75	0.7832	0.7662	0.9638
0.129	45.45	1000	0.1527	0.7306	0.7920	0.7601	0.9619
0.129	50.0	1100	0.1470	0.7429	0.8053	0.7728	0.9638
0.129	54.55	1200	0.1418	0.7552	0.8053	0.7794	0.9657
0.129	59.09	1300	0.1404	0.7657	0.8097	0.7871	0.9667
0.129	63.64	1400	0.1368	0.7741	0.8186	0.7957	0.9695
0.0799	68.18	1500	0.1316	0.7741	0.8186	0.7957	0.9705
0.0799	72.73	1600	0.1301	0.7764	0.8142	0.7948	0.9705
0.0799	77.27	1700	0.1326	0.7699	0.8142	0.7914	0.9695
0.0799	81.82	1800	0.1357	0.7552	0.8053	0.7794	0.9676
0.0799	86.36	1900	0.1304	0.7699	0.8142	0.7914	0.9695
0.0561	90.91	2000	0.1326	0.7699	0.8142	0.7914	0.9695
0.0561	95.45	2100	0.1340	0.7689	0.8097	0.7888	0.9695
0.0561	100.0	2200	0.1371	0.7635	0.8142	0.7880	0.9686
0.0561	104.55	2300	0.1337	0.7764	0.8142	0.7948	0.9705
0.0561	109.09	2400	0.1310	0.7764	0.8142	0.7948	0.9705
0.0451	113.64	2500	0.1353	0.7657	0.8097	0.7871	0.9686
0.0451	118.18	2600	0.1357	0.7657	0.8097	0.7871	0.9686
0.0451	122.73	2700	0.1361	0.7699	0.8142	0.7914	0.9695
0.0451	127.27	2800	0.1358	0.7667	0.8142	0.7897	0.9686
0.0451	131.82	2900	0.1347	0.7699	0.8142	0.7914	0.9695
0.0414	136.36	3000	0.1344	0.7699	0.8142	0.7914	0.9695

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2