Hugging Face has released a new tool called Optimum+ONNX Runtime that allows for faster and easier training of their pre-trained language models. The tool uses the ONNX Runtime to optimize model performance, which can lead to significant speedups in training times. According to Hugging Face, Optimum+ONNX Runtime is up to 2x faster than other popular optimization tools. --- Why it matters: This matters because it can save researchers and engineers time and resources when fine-tuning large language models for specific tasks. Source: https://huggingface.co/blog/optimum-onnxruntime-training