metadata
license: mit
base_model: naver-clova-ix/donut-base
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- bleu
model-index:
- name: donut-base-sroie-v2
results: []
donut-base-sroie-v2
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: 0.1570
- Bleu: 0.0203
- Precisions: [0.9831401475237092, 0.9604966139954854, 0.9343863912515188, 0.9039473684210526]
- Brevity Penalty: 0.0215
- Length Ratio: 0.2067
- Translation Length: 949
- Reference Length: 4592
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- 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 |
|---|---|---|---|---|---|---|---|---|---|
| No log | 0.99 | 62 | 0.2030 | 0.0201 | [0.9789029535864979, 0.9559322033898305, 0.9294403892944039, 0.8985507246376812] | 0.0214 | 0.2064 | 948 | 4592 |
| 0.0922 | 2.0 | 125 | 0.1648 | 0.0207 | [0.9831932773109243, 0.9628796400449944, 0.9382566585956417, 0.90956749672346] | 0.0219 | 0.2073 | 952 | 4592 |
| 0.0922 | 2.99 | 187 | 0.1572 | 0.0202 | [0.9831223628691983, 0.96045197740113, 0.9343065693430657, 0.9038208168642952] | 0.0214 | 0.2064 | 948 | 4592 |
| 0.1224 | 3.97 | 248 | 0.1570 | 0.0203 | [0.9831401475237092, 0.9604966139954854, 0.9343863912515188, 0.9039473684210526] | 0.0215 | 0.2067 | 949 | 4592 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2