donut-base-vishnu3

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: 0.7699

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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.4209	1.0	127	2.1796
2.2026	2.0	254	1.5003
1.3437	3.0	381	1.2493
1.5722	4.0	508	1.1311
0.8412	5.0	635	1.0430
0.4204	6.0	762	0.9765
0.4769	7.0	889	0.9437
0.2852	8.0	1016	0.8951
0.4611	9.0	1143	0.8431
0.2971	10.0	1270	0.8496
0.2414	11.0	1397	0.8118
0.3451	12.0	1524	0.7813
0.1289	13.0	1651	0.7743
0.1248	14.0	1778	0.7769
0.122	15.0	1905	0.7699

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1