donut_experiment_1

This model is a fine-tuned version of naver-clova-ix/donut-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4233
• Bleu: 0.0659
• Precisions: [0.8058455114822547, 0.7440758293838863, 0.7013698630136986, 0.6590909090909091]
• Brevity Penalty: 0.0908
• Length Ratio: 0.2942
• Translation Length: 479
• Reference Length: 1628
• Cer: 0.7576
• Wer: 0.8295

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
0.8942	1.0	253	0.5716	0.0571	[0.7436974789915967, 0.6610978520286396, 0.6104972375690608, 0.5672131147540984]	0.0889	0.2924	476	1628	0.7669	0.8416
0.3794	2.0	506	0.4522	0.0594	[0.770042194092827, 0.697841726618705, 0.6472222222222223, 0.6072607260726073]	0.0876	0.2912	474	1628	0.7642	0.8415
0.3017	3.0	759	0.4154	0.0642	[0.8029350104821803, 0.7357142857142858, 0.6887052341597796, 0.6503267973856209]	0.0895	0.2930	477	1628	0.7577	0.8320
0.222	4.0	1012	0.4233	0.0659	[0.8058455114822547, 0.7440758293838863, 0.7013698630136986, 0.6590909090909091]	0.0908	0.2942	479	1628	0.7576	0.8295

Framework versions

• Transformers 4.40.2
• Pytorch 2.1.0
• Datasets 2.18.0
• Tokenizers 0.19.1