layoutlmv3-final-v6-BI-FINAL

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: 0.4010
• Precision: 0.944
• Recall: 0.9402
• F1: 0.9421
• Accuracy: 0.8895

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	2.56	100	2.1657	0.5961	0.5232	0.5573	0.5357
No log	5.13	200	1.0442	0.8437	0.8353	0.8395	0.7492
No log	7.69	300	0.6743	0.9039	0.8997	0.9018	0.8339
No log	10.26	400	0.5301	0.9219	0.9177	0.9198	0.8553
1.3127	12.82	500	0.4766	0.9340	0.9296	0.9318	0.8673
1.3127	15.38	600	0.4388	0.9414	0.9382	0.9398	0.8762
1.3127	17.95	700	0.4329	0.9311	0.9250	0.9280	0.8705
1.3127	20.51	800	0.4109	0.9385	0.9323	0.9354	0.8850
1.3127	23.08	900	0.4024	0.9434	0.9402	0.9418	0.8901
0.3056	25.64	1000	0.4010	0.944	0.9402	0.9421	0.8895

Framework versions

• Transformers 4.30.0.dev0
• Pytorch 1.8.0+cu101
• Datasets 2.12.0
• Tokenizers 0.13.3