DISC FinLLM

🚀 DISC-FinLLM

DISC-FinLLM is a large financial model based on Baichuan-13B-Chat. It offers professional, intelligent, and comprehensive financial consulting services in various financial scenarios.

Note: Due to the ongoing development of the project, the model weights in this repository may differ from those in our currently deployed demo.

🚀 Quick Start

DISC-FinLLM is a large model in the financial field, developed and open sourced by Fudan University Data Intelligence and Social Computing Laboratory (Fudan-DISC). It's a multi - expert smart financial system with four modules for different financial scenarios.

✨ Features

Financial Consultation

This module enables multiple - round dialogues with users on financial topics in the Chinese financial context and explains relevant financial knowledge. It's based on the financial consulting instructions part of the dataset.

Financial Text Analysis

It helps users complete NLP tasks like information extraction, sentiment analysis, text classification, and text generation on financial texts, trained by financial task instructions in the dataset.

Financial Calculation

This module supports mathematical calculations, including basic ones like interest rates and growth rates, statistical analysis, and financial model calculations such as the Black - Scholes option pricing model and the EDF expected default probability model. It's partially trained from financial computing instructions in the dataset.

Financial Knowledge Retrieval Q&A

Based on financial news, research reports, and related policy documents, it provides users with investment advice, current affairs analysis, and policy interpretation. It's partially trained from retrieval - enhanced instructions in the dataset.

Check our HOME for more information.

📦 Installation

Using through hugging face transformers

>>>import torch
>>>>>>from transformers import AutoModelForCausalLM, AutoTokenizer
>>>from transformers.generation.utils import GenerationConfig
>>>tokenizer = AutoTokenizer.from_pretrained("Go4miii/DISC-FinLLM", use_fast=False, trust_remote_code=True)
>>>model = AutoModelForCausalLM.from_pretrained("Go4miii/DISC-FinLLM", device_map="auto", torch_dtype=torch.float16, trust_remote_code=True)
>>>model.generation_config = GenerationConfig.from_pretrained("Go4miii/DISC-FinLLM")
>>>messages = []
>>>messages.append({"role": "user", "content": "请解释一下什么是银行不良资产？"})
>>>response = model.chat(tokenizer, messages)
>>>print(response)

📚 Documentation

DISC-Fin-SFT Dataset

DISC-FinLLM is based on the high - quality financial dataset DISC-Fin-SFT. We construct and fine - tuned the LoRA instruction on the general - domain Chinese large model Baichuan-13B-Chat. DISC-Fin-SFT contains about 250,000 pieces of data, divided into four sub - datasets: financial consulting instructions, financial task instructions, financial computing instructions, and retrieval - enhanced instructions.

⚠️ Important Note

DISC-FinLLM has problems and shortcomings that cannot be overcome by current large language models. Although it can provide services in the financial field on many tasks and scenarios, the model should be used for user reference only and cannot replace professional financial analysts and financial experts. We hope that users of DISC-FinLLM will be able to critically evaluate the model. We are not responsible for any problems, risks or adverse consequences arising from the use of DISC-FinLLM.

📄 License

The use of the source code in this repository complies with the Apache 2.0 License.

📖 Citation

If our project has been helpful for your research and work, please kindly cite our work as follows:

@misc{yue2023disclawllm,
    title={DISC-LawLLM: Fine-tuning Large Language Models for Intelligent Legal Services}, 
    author={Shengbin Yue and Wei Chen and Siyuan Wang and Bingxuan Li and Chenchen Shen and Shujun Liu and Yuxuan Zhou and Yao Xiao and Song Yun and Xuanjing Huang and Zhongyu Wei},
    year={2023},
    eprint={2309.11325},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}