FinGPT

FinGPT makes financial large language models (FinLLMs) accessible to everyone through FinNLP.

Advantages of FinGPT

1). FinGPT is ideal for finance because of its ability to quickly adapt to changes in the financial landscape. BloombergGPT retrains an LLM using a mixed dataset of finance and general data sources, but this process is costly and time-consuming. FinGPT, on the other hand, can be fine-tuned rapidly, and its adaptation costs are much lower than those of BloombergGPT (estimated at less than $416 per training).

2). FinGPT also provides a more accessible alternative to BloombergGPT. It enables users to access democratized Internet-scale financial data, and to perform timely updates (monthly or weekly updates) using an automated data curation pipeline.

3). FinGPT also employs the unique “Reinforcement Learning from Human Feedback” (RLHF) technology, which is missing in BloombergGPT. RLHF allows an LLM model to learn individual preferences (risk-aversion level, investing habits, personalized robo-advisor, etc.), which is a key factor in ChatGPT and GPT4.

Source: https://github.com/AI4Finance-Foundation/FinGPT