Noureddinesa/Output_LayoutLMv3_v4

Output_LayoutLMv3_v99

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: 0.1581
• Precision: 0.7822
• Recall: 0.7182
• F1: 0.7488
• Accuracy: 0.9619

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-07
• train_batch_size: 3
• eval_batch_size: 3
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 4000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	2.38	100	1.4434	0.0283	0.0636	0.0392	0.6938
No log	4.76	200	0.7802	0.0	0.0	0.0	0.8945
No log	7.14	300	0.5023	0.0	0.0	0.0	0.8962
No log	9.52	400	0.4425	0.0	0.0	0.0	0.8962
0.8848	11.9	500	0.3951	0.0	0.0	0.0	0.8962
0.8848	14.29	600	0.3557	0.0	0.0	0.0	0.8962
0.8848	16.67	700	0.3236	0.0	0.0	0.0	0.8962
0.8848	19.05	800	0.2988	0.2143	0.0273	0.0484	0.8997
0.8848	21.43	900	0.2787	0.4167	0.0909	0.1493	0.9066
0.3328	23.81	1000	0.2623	0.4839	0.1364	0.2128	0.9100
0.3328	26.19	1100	0.2474	0.5238	0.2	0.2895	0.9187
0.3328	28.57	1200	0.2358	0.6038	0.2909	0.3926	0.9308
0.3328	30.95	1300	0.2267	0.6	0.3	0.4	0.9325
0.3328	33.33	1400	0.2172	0.6032	0.3455	0.4393	0.9343
0.2435	35.71	1500	0.2113	0.5821	0.3545	0.4407	0.9343
0.2435	38.1	1600	0.2042	0.5634	0.3636	0.4420	0.9343
0.2435	40.48	1700	0.1981	0.6203	0.4455	0.5185	0.9429
0.2435	42.86	1800	0.1923	0.6628	0.5182	0.5816	0.9446
0.2435	45.24	1900	0.1895	0.6818	0.5455	0.6061	0.9481
0.1971	47.62	2000	0.1846	0.7128	0.6091	0.6569	0.9533
0.1971	50.0	2100	0.1811	0.7526	0.6636	0.7053	0.9585
0.1971	52.38	2200	0.1797	0.7396	0.6455	0.6893	0.9567
0.1971	54.76	2300	0.1755	0.7653	0.6818	0.7212	0.9602
0.1971	57.14	2400	0.1745	0.7526	0.6636	0.7053	0.9585
0.1722	59.52	2500	0.1707	0.7526	0.6636	0.7053	0.9585
0.1722	61.9	2600	0.1672	0.7526	0.6636	0.7053	0.9585
0.1722	64.29	2700	0.1662	0.7677	0.6909	0.7273	0.9602
0.1722	66.67	2800	0.1659	0.7677	0.6909	0.7273	0.9602
0.1722	69.05	2900	0.1650	0.78	0.7091	0.7429	0.9619
0.1558	71.43	3000	0.1633	0.78	0.7091	0.7429	0.9619
0.1558	73.81	3100	0.1613	0.78	0.7091	0.7429	0.9619
0.1558	76.19	3200	0.1605	0.78	0.7091	0.7429	0.9619
0.1558	78.57	3300	0.1600	0.78	0.7091	0.7429	0.9619
0.1558	80.95	3400	0.1594	0.78	0.7091	0.7429	0.9619
0.1461	83.33	3500	0.1588	0.7822	0.7182	0.7488	0.9619
0.1461	85.71	3600	0.1588	0.7822	0.7182	0.7488	0.9619
0.1461	88.1	3700	0.1584	0.7822	0.7182	0.7488	0.9619
0.1461	90.48	3800	0.1583	0.7822	0.7182	0.7488	0.9619
0.1461	92.86	3900	0.1581	0.7822	0.7182	0.7488	0.9619
0.1438	95.24	4000	0.1581	0.7822	0.7182	0.7488	0.9619

Framework versions

• Transformers 4.29.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.13.3