* Fixed typo in multi gpu docs and OLMoE version * Fixed typos in docs for agents, agents advanced, knowledge distillation, and image feature extraction * Fixed incorrect usage of model.image_guided_detection in zero shot object detection docs