Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference

Open in new window