LayoutLMV3_training

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: 0.2281
• Precision: 0.8155
• Recall: 0.86
• F1: 0.8372
• Accuracy: 0.9433

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: 4
• eval_batch_size: 2
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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	0.6667	100	1.6251	0.0512	0.0442	0.0475	0.5825
No log	1.3333	200	1.1656	0.4150	0.4988	0.4531	0.7626
No log	2.0	300	0.7479	0.5971	0.6915	0.6408	0.8657
No log	2.6667	400	0.5020	0.7014	0.7673	0.7329	0.9050
1.1334	3.3333	500	0.3852	0.7695	0.8297	0.7985	0.9251
1.1334	4.0	600	0.3149	0.7837	0.8345	0.8083	0.9327
1.1334	4.6667	700	0.2675	0.8177	0.8588	0.8377	0.9420
1.1334	5.3333	800	0.2460	0.8265	0.8606	0.8432	0.9423
1.1334	6.0	900	0.2363	0.8138	0.8606	0.8365	0.9415
0.3094	6.6667	1000	0.2281	0.8155	0.86	0.8372	0.9433

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1