donut_experiment_5

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: 0.3987
• Bleu: 0.0661
• Precisions: [0.8020833333333334, 0.7375886524822695, 0.6994535519125683, 0.6601941747572816]
• Brevity Penalty: 0.0915
• Length Ratio: 0.2948
• Translation Length: 480
• Reference Length: 1628
• Cer: 0.7576
• Wer: 0.8280

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
0.3274	1.0	253	0.4698	0.0586	[0.7707006369426752, 0.6956521739130435, 0.6582633053221288, 0.62]	0.0857	0.2893	471	1628	0.7660	0.8432
0.2539	2.0	506	0.4198	0.0643	[0.799163179916318, 0.7315914489311164, 0.6868131868131868, 0.6416938110749185]	0.0902	0.2936	478	1628	0.7605	0.8313
0.224	3.0	759	0.3941	0.0658	[0.8075313807531381, 0.7387173396674585, 0.7060439560439561, 0.6710097719869706]	0.0902	0.2936	478	1628	0.7573	0.8283
0.1566	4.0	1012	0.3987	0.0661	[0.8020833333333334, 0.7375886524822695, 0.6994535519125683, 0.6601941747572816]	0.0915	0.2948	480	1628	0.7576	0.8280

Framework versions

• Transformers 4.40.0
• Pytorch 2.1.0
• Datasets 2.18.0
• Tokenizers 0.19.1