donut_experiment_bayesian_trial_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: 0.4541
• Bleu: 0.0661
• Precisions: [0.8029045643153527, 0.731764705882353, 0.6875, 0.639871382636656]
• Brevity Penalty: 0.0928
• Length Ratio: 0.2961
• Translation Length: 482
• Reference Length: 1628
• Cer: 0.7590
• Wer: 0.8315

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: 0.00018010138886762352
• 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: 2
• 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
0.655	1.0	253	0.5770	0.0681	[0.7555110220440882, 0.6719457013574661, 0.625974025974026, 0.5762195121951219]	0.1041	0.3065	499	1628	0.7627	0.8438
0.183	2.0	506	0.4541	0.0661	[0.8029045643153527, 0.731764705882353, 0.6875, 0.639871382636656]	0.0928	0.2961	482	1628	0.7590	0.8315

Framework versions

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