Mapping GPUs to LLMs (and back): A bandwidth-based estimator for local inference(localllm-advisor.com)2 points by apignotti 33 days ago | 0 commentsNo comments yet