---
license: apache-2.0
language: en
library_name: transformers
tags:
  - deberta
  - sequence-classification
  - medicine
  - telerehabilitation
metrics:
  - name: accuracy
    type: accuracy
    value: 0.9986   # test_accuracy
  - name: macro_f1
    type: f1
    value: 0.9987
  - name: balanced_accuracy
    type: balanced_accuracy
    value: 0.9987
  - name: auc_micro
    type: auc
    value: 0.999997
  - name: ap_micro
    type: average_precision
    value: 0.99993
---

# **MeDeBERTa** – v2 (July 2025)

Fine-tuned **microsoft/deberta-v3-xsmall** on 269 874 Q-A pairs (30 intent labels) for the *MeDeBERTa* telerehabilitation question-classification task.

|                                    | Value |
|------------------------------------|-------|
| **Epochs**                         | 20 (best @ epoch 17) |
| **Batch / Grad. Accum.**           | 16 / 4 (eff. 64) |
| **Learning rate**                  | 5 × 10⁻⁵ |
| **Best val. accuracy**        | **0.99855** |
| **Test accuracy**                  | **0.99859** |
| **Macro F1 (test)**           | **0.99867** |
| **Balanced accuracy (test)**       | 0.99868 |
| **Micro AUC**                      | 0.999997 |
| **Micro average precision**        | 0.99993 |
| **Loss (val | test)**              | 0.01371 \| 0.01305 |
| **Hardware**                       | RTX 2080 Ti (11 GB) |

<details>
<summary>Per-class metrics (excerpt)</summary>

| Label | Precision | Recall | F1 | Support |
|-------|-----------|--------|----|---------|
| any_code | 1.000 | 1.000 | 1.000 | 980 |
| contexts | 0.988 | 0.987 | 0.988 | 923 |
| treatment summary | 1.000 | 0.998 | 0.999 | 927 |
| … | … | … | … | … |

Full table: see `classification_report.json` / `classification_report.csv`.
</details>

## Usage

```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification

tok = AutoTokenizer.from_pretrained("malakhovks/MeDeBERTa")
model = AutoModelForSequenceClassification.from_pretrained("malakhovks/MeDeBERTa")

inputs = tok("what are contraindications for TENS?", return_tensors="pt")
pred   = model(**inputs).logits.argmax(-1).item()
print(model.config.id2label[pred])
```

## Changelog
See [CHANGELOG.md](./CHANGELOG.md) for full version history.