| --- |
| base_model: IVN-RIN/medBIT |
| tags: |
| - bert |
| - NER |
| - assertion negation |
| model-index: |
| - name: medbit-assertion-negation |
| results: |
| - task: |
| type: assertion-negation |
| metrics: |
| - name: macro-f1 |
| type: macro-f1 |
| value: 0.946 |
| - name: micro-f1 |
| type: micro-f1 |
| value: 0.946 |
| - name: loss |
| type: loss |
| value: 0.417 |
| language: |
| - it |
| widget: |
| - text: "Il paziente non mostra alcun segno di [entità]." |
| example_title: "Negated" |
| - text: "Il paziente mostra chiari segni di [entità]." |
| example_title: "Affirmed" |
| - text: "Alcuni comportamenti del paziente suggeriscono una ipotetica insorgenza di [entità]. Necessari ulteriori approfondimenti." |
| example_title: "Possible" |
| --- |
| |
| # MedBIT for Clinical Assertion Negation |
|
|
| This model is a fine-tuned version of [IVN-RIN/medBIT-r3-plus](https://huggingface.co/IVN-RIN/medBIT-r3-plus) on a private dataset. |
|
|
| It achieves the following results on the evaluation set: |
| - Loss: 0.417 |
| - Macro-f1: 0.946 |
| - Micro-f1: 0.946 |
|
|
| ## 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-06 |
| - train_batch_size: 4 |
| - eval_batch_size: 4 |
| - seed: 42 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: linear |
| - num_epochs: 21 |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 4.35.2 |
| - Pytorch 2.1.1+cu121 |
| - Datasets 2.15.0 |
| - Tokenizers 0.15.0 |
|
|
