icdar23-entrydetector_plaintext_breaks_indents_left_diff

This model is a fine-tuned version of HueyNemud/das22-10-camembert_pretrained on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0072
  • Ebegin: {'precision': 0.9935508345978755, 'recall': 0.9849567506581421, 'f1': 0.9892351274787535, 'number': 2659}
  • Eend: {'precision': 0.9980857580398163, 'recall': 0.9742152466367713, 'f1': 0.9860060514372163, 'number': 2676}
  • Overall Precision: 0.9958
  • Overall Recall: 0.9796
  • Overall F1: 0.9876
  • Overall Accuracy: 0.9980

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: 0.0001
  • train_batch_size: 2
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 7500

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 0.07 300 0.0309 0.9593 0.9879 0.9734 0.9955
0.161 0.14 600 0.0126 0.9890 0.9911 0.9900 0.9982
0.161 0.21 900 0.0116 0.9730 0.9894 0.9811 0.9971
0.0165 0.29 1200 0.0087 0.9938 0.9918 0.9928 0.9987
0.0119 0.36 1500 0.0093 0.9851 0.9937 0.9894 0.9981
0.0119 0.43 1800 0.0055 0.9942 0.9913 0.9928 0.9987
0.0091 0.5 2100 0.0057 0.9951 0.9904 0.9928 0.9987
0.0091 0.57 2400 0.0058 0.9920 0.9936 0.9928 0.9987
0.0083 0.64 2700 0.0059 0.9896 0.9918 0.9907 0.9983
0.0065 0.72 3000 0.0045 0.9968 0.9917 0.9942 0.9990
0.0065 0.79 3300 0.0047 0.9920 0.9937 0.9929 0.9987
0.0054 0.86 3600 0.0050 0.9945 0.9909 0.9926 0.9987
0.0054 0.93 3900 0.0064 0.9838 0.9968 0.9903 0.9983
0.0056 1.0 4200 0.0046 0.9971 0.9920 0.9946 0.9990
0.0034 1.07 4500 0.0037 0.9959 0.9936 0.9948 0.9990
0.0034 1.14 4800 0.0047 0.9983 0.9900 0.9941 0.9989
0.0035 1.22 5100 0.0043 0.9936 0.9951 0.9944 0.9990
0.0035 1.29 5400 0.0061 0.9892 0.9957 0.9925 0.9986
0.002 1.36 5700 0.0057 0.9898 0.9947 0.9923 0.9986
0.0048 1.43 6000 0.0042 0.9954 0.9933 0.9944 0.9990

Framework versions

  • Transformers 4.26.1
  • Pytorch 1.13.1+cu116
  • Datasets 2.9.0
  • Tokenizers 0.13.2
Downloads last month
11
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.