layoutlmv3-finetuned-cord_100

This model is a fine-tuned version of microsoft/layoutlmv3-base on the cord-layoutlmv3 dataset. It achieves the following results on the evaluation set:

• Loss: 0.3809
• Precision: 0.8837
• Recall: 0.8982
• F1: 0.8909
• Accuracy: 0.9058

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: 1e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 2500

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	0.3125	250	1.4179	0.5786	0.6751	0.6231	0.7037
1.8601	0.625	500	0.9021	0.7458	0.8016	0.7727	0.7988
1.8601	0.9375	750	0.6900	0.8096	0.8338	0.8215	0.8294
0.7675	1.25	1000	0.5915	0.8128	0.8481	0.8300	0.8544
0.7675	1.5625	1250	0.5041	0.8381	0.8638	0.8507	0.8722
0.4979	1.875	1500	0.4669	0.8413	0.8728	0.8567	0.8850
0.4979	2.1875	1750	0.4080	0.8628	0.8847	0.8736	0.8990
0.384	2.5	2000	0.3878	0.8731	0.8907	0.8818	0.9003
0.384	2.8125	2250	0.3880	0.8794	0.8952	0.8872	0.9032
0.3439	3.125	2500	0.3809	0.8837	0.8982	0.8909	0.9058

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0