GenerativeX Releases Fine-Tuning Parameters for a 70 Billion-Parameter Large Language Model (LLM) Specialized for the Financial Industry (Banking, Securities, and Insurance)

GenerativeX Inc., based in Shibuya, Tokyo, is pleased to announce the release of fine-tuning parameters for its proprietary large language model (LLM) developed specifically for the financial industry.

In the financial sector, it is essential to handle various quantitative data accurately, which makes high-performance LLMs indispensable. However, LLMs with high performance, such as ChatGPT, require data to be sent externally, which presents a challenge for the financial industry, where sensitive information must be handled securely.

To address this issue, GenerativeX has developed and released a proprietary LLM (LoRA adapter) that operates on-premises while maintaining high performance in the financial domain. This model is based on Meta’s Llama2, one of the largest open-source models with 70 billion parameters, and has been fine-tuned with knowledge specific to the financial industry.