Edit model card

donut-base-sroie-test-050824

This model is a fine-tuned version of naver-clova-ix/donut-base on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5823
  • Bleu: 0.0595
  • Precisions: [0.7332123411978222, 0.6290983606557377, 0.5741176470588235, 0.5248618784530387]
  • Brevity Penalty: 0.0974
  • Length Ratio: 0.3004
  • Translation Length: 551
  • Reference Length: 1834
  • Cer: 0.7739
  • Wer: 0.8665

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: 2e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 2
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 4
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Bleu Precisions Brevity Penalty Length Ratio Translation Length Reference Length Cer Wer
3.716 1.0 250 1.7346 0.0080 [0.3276089828269485, 0.05475504322766571, 0.01901743264659271, 0.0035211267605633804] 0.2411 0.4128 757 1834 0.8647 0.9639
1.3396 2.0 500 0.8150 0.0346 [0.6405353728489483, 0.46304347826086956, 0.3702770780856423, 0.2964071856287425] 0.0815 0.2852 523 1834 0.7877 0.8990
0.8804 3.0 750 0.6424 0.0586 [0.7463235294117647, 0.6486486486486487, 0.5933014354066986, 0.5408450704225352] 0.0934 0.2966 544 1834 0.7674 0.8638
0.6436 4.0 1000 0.5823 0.0595 [0.7332123411978222, 0.6290983606557377, 0.5741176470588235, 0.5248618784530387] 0.0974 0.3004 551 1834 0.7739 0.8665

Framework versions

  • Transformers 4.41.0.dev0
  • Pytorch 2.1.0
  • Datasets 2.19.1
  • Tokenizers 0.19.1
Downloads last month
0
Safetensors
Model size
202M params
Tensor type
I64
·
F32
·
Unable to determine this model’s pipeline type. Check the docs .

Finetuned from