molbal/trading-llm-experiment

Experimenting whether finetuning an LLM with historical data from an exchange can teach it to predict the price. (Spoiler alert: no it can't)

This repository is not active