metadata
license: mit
base_model: philschmid/donut-base-sroie
tags:
- generated_from_trainer
datasets:
- imagefolder
model-index:
- name: donut-stnk-v5_1-left-sroie-based
results: []
donut-stnk-v5_1-left-sroie-based
This model is a fine-tuned version of philschmid/donut-base-sroie on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0801
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: 2
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.2063 | 1.0 | 224 | 0.0979 |
| 0.1059 | 2.0 | 448 | 0.0835 |
| 0.0698 | 3.0 | 672 | 0.0801 |
| 0.0445 | 4.0 | 896 | 0.0817 |
| 0.0376 | 5.0 | 1120 | 0.0848 |
| 0.0224 | 6.0 | 1344 | 0.0951 |
Framework versions
- Transformers 4.37.2
- Pytorch 1.13.1+cu117
- Datasets 2.16.1
- Tokenizers 0.15.1