File size: 915 Bytes
b786f3a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 | # Action Requirement Classifier
This repository contains a lightweight text classifier for predicting whether a message is:
- `action_required`
- `informational`
- `optional_action`
## What’s inside
- `classifier.joblib`: a scikit-learn classifier trained on SentenceTransformer embeddings
- `encoder/`: the SentenceTransformer encoder saved via `SentenceTransformer.save()`
- `metadata.json`: bundle metadata (model selection metric, encoder name, etc.)
## Quickstart (Python)
```python
import joblib
from sentence_transformers import SentenceTransformer
encoder = SentenceTransformer("./encoder")
clf = joblib.load("classifier.joblib")
texts = ["Please upload a clearer photo of your ID."]
emb = encoder.encode(texts)
pred = clf.predict(emb)
print(pred[0])
```
## Notes
- The classifier expects embeddings from the included `encoder/` directory.
- If you move files around, update paths accordingly.
|