DunnBC22's picture
Update README.md
4ca73d1
|
raw
history blame
2.98 kB
metadata
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - recall
  - precision
model-index:
  - name: dit-base-Document_Classification-RVL_CDIP
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: data
          split: train
          args: data
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.976678084687705
language:
  - en

dit-base-Document_Classification-RVL_CDIP

This model is a fine-tuned version of microsoft/dit-base.

It achieves the following results on the evaluation set:

  • Loss: 0.0786
  • Accuracy: 0.9767
  • Weighted f1: 0.9768
  • Micro f1: 0.9767
  • Macro f1: 0.9154
  • Weighted recall: 0.9767
  • Micro recall: 0.9767
  • Macro recall: 0.9019
  • Weighted precision: 0.9771
  • Micro precision: 0.9767
  • Macro precision: 0.9314

Model description

For more information on how it was created, check out the following link: https://github.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/blob/main/Document%20AI/Multiclass%20Classification/Document%20Classification%20-%20RVL-CDIP/Document%20Classification%20-%20RVL-CDIP.ipynb

Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

Training and evaluation data

Dataset Source: https://www.kaggle.com/datasets/achrafbribiche/document-classification

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss Accuracy Weighted f1 Micro f1 Macro f1 Weighted recall Micro recall Macro recall Weighted precision Micro precision Macro precision
0.1535 1.0 208 0.1126 0.9622 0.9597 0.9622 0.5711 0.9622 0.9622 0.5925 0.9577 0.9622 0.5531
0.1195 2.0 416 0.0843 0.9738 0.9736 0.9738 0.8502 0.9738 0.9738 0.8037 0.9741 0.9738 0.9287
0.0979 3.0 624 0.0786 0.9767 0.9768 0.9767 0.9154 0.9767 0.9767 0.9019 0.9771 0.9767 0.9314

Framework versions

  • Transformers 4.28.1
  • Pytorch 2.0.0
  • Datasets 2.11.0
  • Tokenizers 0.13.3