A reproducible benchmark of Text Generation Inference and Transformers as of April 2024 on AMD Instinct MI250 and MI300 - View it on GitHub