🤖 Intent Detection using Fine-Tuned BERT

This project utilizes a fine-tuned BERT model (bert-base-uncased) for intent classification tasks. It is an encoder-only transformer designed to detect user intents from text inputs (e.g., chatbot queries) and classify them into predefined categories such as banking, travel, finance, and more.

The model is trained on the CLINC150 (clinc_oos) dataset and evaluated using accuracy as the primary metric.

📊 Dataset --> CLINC150

The project uses the CLINC150 dataset, a benchmark dataset for intent classification in task-oriented dialogue systems.

🧾 Dataset Overview

Total intents: 150 unique user intents
Domains: 10 real-world domains (e.g., banking, travel, weather, small talk)
Examples: ~22,500 utterances
Language: English
Out-of-scope (OOS): Includes OOS examples to test robustness

📦 Source

Official repo: clinc/oos-eval
Hugging Face: clinc_oos

🚀 Example

Request: "I want to book a flight"

Response: "book_flight"

Downloads last month: 4

Safetensors

Model size

0.1B params

Tensor type

F32

Model tree for SaherMuhamed/bert-intention-classifier

Base model

google-bert/bert-base-uncased

Finetuned

(6725)

this model

SaherMuhamed
/

bert-intention-classifier