Bringing PyTorch Monarch to AMD GPUs

Featured projects Training state-of-the-art large language models (LLMs) with billions of parameters requires distributed training across hundreds or thousands of GPUs. At this scale, hardware failure...