layoutlmv3-test

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

• Loss: 0.8036
• Precision: 0.8973
• Recall: 0.9200
• F1: 0.9085
• Accuracy: 0.8481

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	5.26	100	0.5115	0.8071	0.8624	0.8338	0.8407
No log	10.53	200	0.4661	0.8730	0.9086	0.8905	0.8546
No log	15.79	300	0.5613	0.8914	0.9091	0.9001	0.8552
No log	21.05	400	0.6767	0.8937	0.8982	0.8959	0.8507
0.3022	26.32	500	0.7020	0.8935	0.9165	0.9049	0.8626
0.3022	31.58	600	0.7108	0.9040	0.9220	0.9129	0.8591
0.3022	36.84	700	0.7378	0.9049	0.9175	0.9112	0.8517
0.3022	42.11	800	0.7892	0.9026	0.9210	0.9117	0.8537
0.3022	47.37	900	0.8133	0.8995	0.9205	0.9099	0.8490
0.0223	52.63	1000	0.8036	0.8973	0.9200	0.9085	0.8481

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1