DistiLLM: Towards Streamlined Distillation for Large Language Models

Open in new window