donut_experiment_bayesian_trial_18

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: 0.5643
• Bleu: 0.0698
• Precisions: [0.8340248962655602, 0.7741176470588236, 0.7309782608695652, 0.6784565916398714]
• Brevity Penalty: 0.0928
• Length Ratio: 0.2961
• Translation Length: 482
• Reference Length: 1628
• Cer: 0.7496
• Wer: 0.8244

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: 1.7803961202565393e-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: 3
• 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
0.0287	1.0	253	0.5097	0.0722	[0.8374485596707819, 0.7762237762237763, 0.7338709677419355, 0.6888888888888889]	0.0954	0.2985	486	1628	0.7506	0.8208
0.0159	2.0	506	0.5583	0.0697	[0.8319502074688797, 0.7741176470588236, 0.7282608695652174, 0.6784565916398714]	0.0928	0.2961	482	1628	0.7496	0.8232
0.0118	3.0	759	0.5643	0.0698	[0.8340248962655602, 0.7741176470588236, 0.7309782608695652, 0.6784565916398714]	0.0928	0.2961	482	1628	0.7496	0.8244

Framework versions

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