AttnCache: Accelerating Self-Attention Inference for LLM Prefill via Attention Cache

Open in new window