This folder contains examples of running Speculative-Decoding Examples with IPEX-LLM on Intel GPU:

• Self-Speculation: running BF16 inference for Huggingface Transformer model with self-speculative decoding with IPEX-LLM on Intel GPUs
• EAGLE: running speculative sampling using EAGLE (Extrapolation Algorithm for Greater Language-model Efficiency) with IPEX-LLM on Intel GPUs