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: 0.3404
• Precision: 0.9296
• Recall: 0.9386
• F1: 0.9341
• Accuracy: 0.9304

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: 5
• eval_batch_size: 5
• 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	4.17	250	0.9987	0.7470	0.7957	0.7706	0.8043
1.3632	8.33	500	0.5299	0.8641	0.8855	0.8747	0.8829
1.3632	12.5	750	0.3861	0.8853	0.9124	0.8986	0.9126
0.3151	16.67	1000	0.3392	0.9154	0.9311	0.9232	0.9321
0.3151	20.83	1250	0.3382	0.9247	0.9371	0.9309	0.9308
0.1265	25.0	1500	0.3364	0.9225	0.9356	0.9290	0.9300
0.1265	29.17	1750	0.3333	0.9304	0.9401	0.9352	0.9321
0.0716	33.33	2000	0.3381	0.9296	0.9394	0.9345	0.9312
0.0716	37.5	2250	0.3474	0.9290	0.9409	0.9349	0.9321
0.0525	41.67	2500	0.3404	0.9296	0.9386	0.9341	0.9304

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.11.0
• Tokenizers 0.13.2