donut-base-sroie-metrics-combined-new

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: 0.4671
• Bleu: 0.0662
• Precisions: [0.785140562248996, 0.6825396825396826, 0.6197916666666666, 0.5626911314984709]
• Brevity Penalty: 0.1007
• Length Ratio: 0.3035
• Translation Length: 498
• Reference Length: 1641
• Cer: 0.7528
• Wer: 0.8385

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
3.6559	1.0	253	1.5613	0.0007	[0.5056179775280899, 0.1943127962085308, 0.07692307692307693, 0.02830188679245283]	0.0058	0.1627	267	1641	0.8768	0.9436
1.2493	2.0	506	0.6697	0.0409	[0.6560509554140127, 0.5048309178743962, 0.4481792717086835, 0.39]	0.0834	0.2870	471	1641	0.7766	0.8837
0.9257	3.0	759	0.5168	0.0594	[0.75, 0.6275862068965518, 0.5714285714285714, 0.5264797507788161]	0.0968	0.2998	492	1641	0.7570	0.8499
0.6416	4.0	1012	0.4671	0.0662	[0.785140562248996, 0.6825396825396826, 0.6197916666666666, 0.5626911314984709]	0.1007	0.3035	498	1641	0.7528	0.8385

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.1.0
• Datasets 2.19.1
• Tokenizers 0.19.1