layoutmlv3_thursday_sep7_v5

This model is a fine-tuned version of microsoft/layoutlmv3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1416
• Precision: 0.5517
• Recall: 0.9412
• F1: 0.6957
• Accuracy: 0.9822

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	8.33	100	0.3243	0.5556	0.8824	0.6818	0.9485
No log	16.67	200	0.1584	0.6087	0.8235	0.7	0.9734
No log	25.0	300	0.1682	0.5517	0.9412	0.6957	0.9769
No log	33.33	400	0.1773	0.4545	0.8824	0.6	0.9734
0.2633	41.67	500	0.1631	0.4375	0.8235	0.5714	0.9751
0.2633	50.0	600	0.1526	0.5517	0.9412	0.6957	0.9769
0.2633	58.33	700	0.1430	0.5517	0.9412	0.6957	0.9840
0.2633	66.67	800	0.1497	0.5517	0.9412	0.6957	0.9822
0.2633	75.0	900	0.1418	0.5517	0.9412	0.6957	0.9805
0.0111	83.33	1000	0.1416	0.5517	0.9412	0.6957	0.9822

Framework versions

• Transformers 4.34.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3