donut_synDB_test_new
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: 0.0795
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: 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 |
---|---|---|---|
0.7616 | 0.21 | 50 | 0.6341 |
0.3537 | 0.31 | 75 | 0.3127 |
0.2538 | 0.42 | 100 | 0.2624 |
0.1609 | 0.52 | 125 | 0.2998 |
0.1056 | 0.62 | 150 | 0.1088 |
0.0803 | 0.73 | 175 | 0.1888 |
0.0678 | 0.83 | 200 | 0.1151 |
0.0619 | 0.94 | 225 | 0.1307 |
0.0379 | 1.04 | 250 | 0.1469 |
0.057 | 1.15 | 275 | 0.1348 |
0.035 | 1.25 | 300 | 0.1238 |
0.0438 | 1.35 | 325 | 0.1593 |
0.0412 | 1.46 | 350 | 0.1329 |
0.0255 | 1.56 | 375 | 0.1216 |
0.0556 | 1.67 | 400 | 0.1018 |
0.0273 | 1.77 | 425 | 0.1704 |
0.0273 | 1.88 | 450 | 0.0689 |
0.0216 | 1.98 | 475 | 0.0512 |
0.0143 | 2.08 | 500 | 0.0753 |
0.006 | 2.19 | 525 | 0.0763 |
0.0178 | 2.29 | 550 | 0.0724 |
0.0165 | 2.4 | 575 | 0.0738 |
0.0204 | 2.5 | 600 | 0.0777 |
0.0112 | 2.6 | 625 | 0.0759 |
0.0087 | 2.71 | 650 | 0.1009 |
0.0158 | 2.81 | 675 | 0.0812 |
0.0128 | 2.92 | 700 | 0.0954 |
0.0272 | 3.02 | 725 | 0.1064 |
0.0037 | 3.12 | 750 | 0.1140 |
0.024 | 3.23 | 775 | 0.1509 |
0.0082 | 3.33 | 800 | 0.1103 |
0.023 | 3.44 | 825 | 0.0999 |
0.0104 | 3.54 | 850 | 0.1040 |
0.0063 | 3.65 | 875 | 0.0996 |
0.013 | 3.75 | 900 | 0.0852 |
0.0129 | 3.85 | 925 | 0.0734 |
0.0084 | 3.96 | 950 | 0.0732 |
0.0039 | 4.06 | 975 | 0.0795 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 12
Unable to determine this model’s pipeline type. Check the
docs
.