From Words to Watts: Benchmarking the Energy Costs of Large Language Model Inference

Open in new window