layoutlmv3_document_classification

This model is a fine-tuned version of microsoft/layoutlmv3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8183
• Accuracy: 0.8340
• F1: 0.8221

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
• lr_scheduler_warmup_steps: 500
• num_epochs: 7

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.6545	0.4274	100	0.9149	0.8233	0.8041
0.6313	0.8547	200	0.9024	0.8276	0.8085
0.5714	1.2821	300	0.9112	0.8223	0.8067
0.5408	1.7094	400	0.8816	0.8298	0.8093
0.482	2.1368	500	0.9015	0.8244	0.8081
0.4546	2.5641	600	0.8779	0.8287	0.8180
0.4718	2.9915	700	0.8879	0.8212	0.8056
0.4302	3.4188	800	0.8562	0.8276	0.8155
0.5039	3.8462	900	0.8382	0.8330	0.8226
0.4644	4.2735	1000	0.8455	0.8308	0.8200
0.4411	4.7009	1100	0.8461	0.8308	0.8228
0.4007	5.1282	1200	0.8304	0.8308	0.8200
0.4023	5.5556	1300	0.8370	0.8330	0.8242
0.3756	5.9829	1400	0.8193	0.8405	0.8286
0.3592	6.4103	1500	0.8185	0.8394	0.8282
0.3429	6.8376	1600	0.8183	0.8340	0.8221

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0