Together AIBlog12 FebIntroducing Dedicated Container Inference: Delivering 2.6x faster inference for custom AI models