|
--- |
|
license: apache-2.0 |
|
tags: |
|
- setfit |
|
- sentence-transformers |
|
- text-classification |
|
pipeline_tag: text-classification |
|
library_name: sentence-transformers |
|
--- |
|
|
|
# splore/navigational-vs-transactional-vs-informational-classifier |
|
|
|
Query classification model trained using Metaflow. |
|
|
|
## Query types |
|
1. Navigational |
|
2. Transactional |
|
3. Informational |
|
|
|
## Usage |
|
|
|
```python |
|
from setfit import SetFitModel |
|
|
|
model = SetFitModel.from_pretrained("splore/navigational-vs-transactional-vs-informational-classifier") |
|
predictions = model.predict(["What type of query is this?"]) |
|
probabilities = model.predict_proba(["Predictions with probabilities"]) |
|
``` |
|
|
|
## Evaluation |
|
|
|
Tests have only been done on a small dataset of 40 manually created queries. |
|
|
|
|metric |value| |
|
|-----------|-----| |
|
|accuracy |0.942| |
|
|exact match|0.875| |
|
|
|
 |
|
|
|
 |
