|
|
--- |
|
|
license: mit |
|
|
tags: |
|
|
- finance |
|
|
- financial-qa |
|
|
- finetuned-llm |
|
|
- phi3 |
|
|
- phi-3-mini |
|
|
- fintech |
|
|
- accounting |
|
|
- banking |
|
|
- investment |
|
|
- risk-analysis |
|
|
- lora |
|
|
- merged-model |
|
|
language: |
|
|
- en |
|
|
pipeline_tag: text-generation |
|
|
base_model: microsoft/phi-3-mini-4k-instruct |
|
|
model_name: finance_llm_full |
|
|
model_creator: devAnurag |
|
|
pretty_name: Finance LLM Full |
|
|
--- |
|
|
|
|
|
# 💎 Finance LLM Full — Next-Gen Financial Intelligence Model |
|
|
<p align="center"> |
|
|
<img src="finanace_llm.png" alt="Finance LLM Logo" width="180" style="border-radius:12px;"/> |
|
|
</p> |
|
|
|
|
|
<h1 align="center">💼 Finance LLM Full</h1> |
|
|
|
|
|
<p align="center"> |
|
|
<b>A next-generation Financial Intelligence Model<br> |
|
|
Fine-Tuned, Merged & Optimized for Real-World Finance</b> |
|
|
</p> |
|
|
|
|
|
<p align="center"> |
|
|
<a href="https://huggingface.co/devAnurag/finance_llm_full"> |
|
|
<img src="https://img.shields.io/badge/HuggingFace-Model-yellow?logo=huggingface" /> |
|
|
</a> |
|
|
<img src="https://img.shields.io/badge/Model%20Type-Merged%20LoRA-blue" /> |
|
|
<img src="https://img.shields.io/badge/Base%20Model-Phi3%20Mini%204K-green" /> |
|
|
<img src="https://img.shields.io/badge/Domain-Finance%20%26%20Business-purple" /> |
|
|
</p> |
|
|
|
|
|
--- |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
**Finance LLM Full** is a high-performance, fully merged financial Large Language Model (LLM) |
|
|
designed to deliver **crystal-clear, accurate, and structured financial reasoning**. |
|
|
|
|
|
It is trained using **LoRA fine-tuning** on top of **Phi-3 Mini 4K Instruct**, and later |
|
|
**merged into a single standalone model** for seamless deployment. |
|
|
|
|
|
This model specializes in **Finance, Accounting, Banking, Investment, Stock Markets, and Business Analysis** — |
|
|
making it ideal for **FinTech products, AI advisors, investment copilots, and enterprise bots**. |
|
|
|
|
|
--- |
|
|
|
|
|
# ⚡ Why Finance LLM Full is Special |
|
|
|
|
|
### 🔹 1. Purpose-Built For Finance |
|
|
Unlike general LLMs, this model deeply understands: |
|
|
- Balance Sheet Interpretation |
|
|
- Profit & Loss Breakdown |
|
|
- Cashflow Logic |
|
|
- EBITDA / EPS / ROE / DCF |
|
|
- Risk & Return Analysis |
|
|
- Banking, Loans, Limits, Credit Rules |
|
|
- Valuation Basics |
|
|
- Investment & Portfolio Concepts |
|
|
|
|
|
### 🔹 2. Merged Model → One File, Zero Hassle |
|
|
✔ No LoRA needed |
|
|
✔ No adapter loading |
|
|
✔ Direct plug-and-play |
|
|
✔ Works on CPU / GPU / Colab / Docker |
|
|
|
|
|
### 🔹 3. Small Model → Big Capability |
|
|
Powered by **Phi-3 Mini**, optimized for: |
|
|
- Low latency |
|
|
- Low VRAM/RAM usage |
|
|
- Clean, structured answers |
|
|
- High domain accuracy |
|
|
|
|
|
--- |
|
|
|
|
|
# 🧪 Quick Start (Copy & Run) |
|
|
|
|
|
```python |
|
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
|
import torch |
|
|
|
|
|
model_id = "devAnurag/finance_llm_full" |
|
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained(model_id) |
|
|
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.bfloat16) |
|
|
|
|
|
prompt = "Explain the difference between EBITDA and Net Profit." |
|
|
inputs = tokenizer(prompt, return_tensors="pt") |
|
|
|
|
|
outputs = model.generate(**inputs, max_new_tokens=150) |
|
|
print(tokenizer.decode(outputs[0], skip_special_tokens=True)) |
|
|
|
