Hugging Face's logo

Spaces:
mobadara
/
finbert-sentiment-api
Running

App Files Community
finbert-sentiment-api
File size: 2,773 Bytes
9ecc3d2
 
 
 
 
 
 
 
 
 
11dfce2
 
 
 
 
 
 
 
c66b19a
0f7b2eb
11dfce2
 
 
 
 
 
 
 
 
 
5738f9e
11dfce2
 
 
 
 
 
 
 
 
5738f9e
 
11dfce2
 
 
 
 
 
9ecc3d2
11dfce2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9ecc3d2
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
---
title: FinBERT Sentiment Analyzer API
emoji: 📊
colorFrom: blue
colorTo: gray
sdk: docker
pinned: false
---

# 📈 **FinBERT: Real-Time Financial Sentiment Analysis**

Machine Learning Pipeline, which analyses news headlines about finance and forecasts sentiments (Bullish, Bearish, and Neutral).

The project will train a BERT-based model with PyTorch, implement API prediction requests using FastAPI, and display visualization results on a React-Bootstrap web interface.

### 🔗 Project Links
* **Live Application:** [Link](https://portfolio-frontend-livid.vercel.app/projects/sentiment-analysis-with-bert)
* **Frontend Code:** [Github](https://github.com/mobadara/portfolio-frontend/blob/main/src/pages/projects/FinBERT.jsx)
* **Project Blog Post:** [Link](https://mobadara.medium.com/75740300902b?source=friends_link&sk=e145c95c5c6b232c3b3ba27eaf0b41fb)
* **Model Weights (Optional):** Hugging Face [mobadara/finbert-finetuned](https://huggingface.co/mobadara/finbert-finetuned)

---

## 🏗️ System Architecture

* **Machine Learning:** PyTorch, Hugging Face `transformers`, Financial PhraseBank Dataset
* **Backend API:** Python, FastAPI, Uvicorn
* **Frontend UI:** React, React-Bootstrap
* **Database:** PostgreSQL (Neon/Supabase) via SQLAlchemy

---

## 🧠 The Machine Learning Pipeline

The core of this application is a fine-tuned NLP model.
1. **Base Model:** `ProsusAI/finbert`
2. **Fine-tuning:** Conducted in Google Colab using a T4 GPU. 
3. **Training Data:** The Kaggle Financial PhraseBank dataset.
*(Check the `/notebooks` directory to see the complete PyTorch training loop, tokenization process, and evaluation metrics).*

---

## 🚀 How to Run Locally

### 1. Clone the Repository

copy the command below and run it in your favourite terminal.
```bash
git clone https://github.com/mobadara/finbert-sentiment-analyzer-api &&
cd finbert-sentiment-analyzer-api.git
```

### 2. Create a virtual environment
```
python -m venv venv
```

### 3. Activate the virtual environment
#### a. On Linux/Mac
```bash
source venv/bin/activate
```

#### b. On Windows
```bash
venv\Scripts\activate
```

### 4. Install Dependencies
```bash
pip install -r requirements.txt
```

### 5. Start the server
```bash
uvicorn app.main:app --reload
```

## 👨‍💻 Author:
**Muyiwa J. Obadara**

[![LinkedIn](https://img.shields.io/badge/LinkedIn-%230077B5.svg?style=for-the-badge&logo=linkedin&logoColor=white)](https://linkedin.com/in/obadara-m)
[![X / Twitter](https://img.shields.io/badge/X-%23000000.svg?style=for-the-badge&logo=X&logoColor=white)](https://twitter.com/m_obadara)
[![Portfolio](https://img.shields.io/badge/Portfolio-Navy_Blue?style=for-the-badge&logo=google-chrome&logoColor=white&color=navy)](https://portfolio-frontend-livid.vercel.app)