layoutlmv3-finetuned-funsd

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: 1.1059
• Precision: 0.7823
• Recall: 0.8281
• F1: 0.8045
• Accuracy: 0.8135

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	10.0	100	0.7158	0.6964	0.7690	0.7309	0.7833
No log	20.0	200	0.7508	0.7420	0.8058	0.7726	0.8076
No log	30.0	300	0.8085	0.7678	0.8097	0.7882	0.8078
No log	40.0	400	0.9083	0.7689	0.8147	0.7911	0.8061
0.3419	50.0	500	0.9294	0.7840	0.8296	0.8062	0.8128
0.3419	60.0	600	1.0036	0.7807	0.8366	0.8077	0.8178
0.3419	70.0	700	1.0983	0.7727	0.8241	0.7976	0.8044
0.3419	80.0	800	1.1092	0.7921	0.8251	0.8083	0.8064
0.3419	90.0	900	1.1010	0.7780	0.8306	0.8035	0.8133
0.0261	100.0	1000	1.1059	0.7823	0.8281	0.8045	0.8135

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1