Optimizing LLaMA-3 8B for Bitcoin Price Forecasting via PRICE
September 7, 2026
The PRICE framework optimizes a 4-bit quantized LLaMA-3 8B for cryptocurrency forecasting using LoRA, recursive multi-step inference, and integer-rounded numerical representation. The study evaluates how these specific adaptation choices jointly impact short-term price prediction accuracy.
HOW THIS AFFECTS YOU
●
builderYou can improve time-series LLM performance by combining LoRA with integer-rounded numerical representations.
●
researcherYou can study the intersection of quantization and decoding strategies in non-stationary financial environments.