layoutlmv3-large-model2ab-telephone-consultation-discharge

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

• Loss: 0.2897
• Accuracy: 0.9334
• Macro Precision: 0.9333
• Macro Recall: 0.9365
• Macro F1: 0.9344
• Weighted F1: 0.9332
• Precision Telephone Encounter: 0.9283
• Recall Telephone Encounter: 0.9510
• F1 Telephone Encounter: 0.9395
• Precision Consultation Note: 0.9439
• Recall Consultation Note: 0.8899
• F1 Consultation Note: 0.9161
• Precision Discharge Report: 0.9277
• Recall Discharge Report: 0.9686
• F1 Discharge Report: 0.9477

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: 2e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAFACTOR and the args are: No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Macro Precision	Macro Recall	Macro F1	Weighted F1	Precision Telephone Encounter	Recall Telephone Encounter	F1 Telephone Encounter	Precision Consultation Note	Recall Consultation Note	F1 Consultation Note	Precision Discharge Report	Recall Discharge Report	F1 Discharge Report
0.1719	1.4216	500	0.2575	0.9303	0.9332	0.9330	0.9331	0.9303	0.9187	0.9224	0.9206	0.9248	0.9207	0.9227	0.9560	0.9560	0.9560
0.1268	2.8432	1000	0.2906	0.9334	0.9333	0.9365	0.9344	0.9332	0.9283	0.9510	0.9395	0.9439	0.8899	0.9161	0.9277	0.9686	0.9477

Framework versions

• Transformers 4.57.6
• Pytorch 2.10.0+cu128
• Tokenizers 0.22.2