arquivo-layoutxml-model

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

• Loss: 0.2997
• Ategory Precision: 0.8719
• Ategory Recall: 0.8491
• Ategory F1: 0.8603
• Ategory Number: 497
• Itle Precision: 0.8745
• Itle Recall: 0.8971
• Itle F1: 0.8857
• Itle Number: 2508
• One Precision: 0.8855
• One Recall: 0.8855
• One F1: 0.8855
• One Number: 2951
• Ubtitle Precision: 0.9494
• Ubtitle Recall: 0.9774
• Ubtitle F1: 0.9632
• Ubtitle Number: 23695
• Overall Precision: 0.9356
• Overall Recall: 0.9593
• Overall F1: 0.9473
• Overall Accuracy: 0.9629

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: 5e-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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3