layoutlmv3-ap7_2_stitched

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: 0.0223
• Precision: 0.4286
• Recall: 0.5769
• F1: 0.4918
• Accuracy: 0.9956

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: 4
• 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_ratio: 0.1
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.5122	13.8889	250	0.0918	0.0	0.0	0.0	0.9862
0.04	27.7778	500	0.0346	0.1786	0.1923	0.1852	0.9910
0.0218	41.6667	750	0.0262	0.4688	0.5769	0.5172	0.9948
0.0159	55.5556	1000	0.0223	0.4286	0.5769	0.4918	0.9956

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1