donut-base-beans

This model is a fine-tuned version of microsoft/dit-base-finetuned-rvlcdip on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1870
• F1: 0.3826

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: 0.0005
• train_batch_size: 128
• eval_batch_size: 128
• seed: 1337
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• training_steps: 30000

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	0.2524	500	0.1623	0.3040
0.1656	0.5048	1000	0.4307	nan
0.1656	0.7572	1500	0.1493	nan
0.14	1.0096	2000	0.1870	0.3826
0.14	1.2620	2500	0.1076	nan
0.1285	1.5144	3000	0.0896	0.3812
0.1285	1.7668	3500	0.1405	nan
0.123	2.0192	4000	0.2061	nan
0.123	2.2716	4500	0.2283	0.2395

Framework versions

• Transformers 4.43.0.dev0
• Pytorch 2.1.2+cu118
• Datasets 2.20.0
• Tokenizers 0.19.1