layoutlmv3-finetuned-funsd

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

• Loss: 1.1951
• Precision: 0.9104
• Recall: 0.9086
• F1: 0.9095
• Accuracy: 0.8530

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: 5
• eval_batch_size: 5
• 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	3.3333	100	0.8172	0.8957	0.9046	0.9001	0.8418
No log	6.6667	200	0.8379	0.8870	0.9160	0.9013	0.8381
No log	10.0	300	0.9611	0.8887	0.9041	0.8963	0.8328
No log	13.3333	400	0.9324	0.8888	0.9091	0.8988	0.8438
0.0651	16.6667	500	0.9475	0.8928	0.9185	0.9055	0.8511
0.0651	20.0	600	1.1234	0.8834	0.9031	0.8931	0.8343
0.0651	23.3333	700	1.1130	0.8921	0.8957	0.8939	0.8254
0.0651	26.6667	800	1.0760	0.8931	0.9175	0.9052	0.8416
0.0651	30.0	900	1.1777	0.8894	0.9031	0.8962	0.8336
0.0115	33.3333	1000	1.2102	0.9025	0.9101	0.9063	0.8387
0.0115	36.6667	1100	1.1602	0.9012	0.9111	0.9061	0.8467
0.0115	40.0	1200	1.1819	0.9011	0.9101	0.9056	0.8433
0.0115	43.3333	1300	1.2095	0.9051	0.9051	0.9051	0.8452
0.0115	46.6667	1400	1.1687	0.9064	0.9185	0.9124	0.8570
0.0031	50.0	1500	1.1951	0.9104	0.9086	0.9095	0.8530
0.0031	53.3333	1600	1.1967	0.9041	0.9131	0.9086	0.8530
0.0031	56.6667	1700	1.1989	0.9015	0.9091	0.9053	0.8531
0.0031	60.0	1800	1.1973	0.9000	0.9126	0.9063	0.8549
0.0031	63.3333	1900	1.2135	0.9015	0.9096	0.9055	0.8490
0.0012	66.6667	2000	1.2210	0.9015	0.9091	0.9053	0.8469

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.19.1