|
|
--- |
|
|
language: en |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- text-classification |
|
|
- banking |
|
|
- intent-detection |
|
|
- transformers |
|
|
library_name: transformers |
|
|
pipeline_tag: text-classification |
|
|
model_type: bert |
|
|
metrics: |
|
|
- accuracy |
|
|
- recall |
|
|
- precision |
|
|
base_model: |
|
|
- google-bert/bert-base-uncased |
|
|
--- |
|
|
|
|
|
# Question Classification Model for Bank Queries |
|
|
|
|
|
This model is fine-tuned specifically for banking-related queries to classify whether a user intends to perform a **transaction** or not. |
|
|
|
|
|
## 🧠 Use Case |
|
|
|
|
|
Given a text input (a user question or statement), the model returns: |
|
|
- `"True"`: if the query is a **question** |
|
|
- `"False"`: otherwise |
|
|
|
|
|
--- |
|
|
|
|
|
## 🔧 How to Use |
|
|
|
|
|
You can use this model directly with the Hugging Face `transformers` pipeline: |
|
|
|
|
|
```python |
|
|
from transformers import pipeline |
|
|
|
|
|
hf_model = "pankaj1881/question-classification" |
|
|
|
|
|
classifier = pipeline("text-classification", model=hf_model) |
|
|
|
|
|
query = "I want to transfer 500 dollars to my friend" |
|
|
result = classifier(query) |
|
|
|
|
|
print(result) |
|
|
# Output example: [{'label': 'False', 'score': 0.8767889142036438}] i.e it's not a question. |
