OpenChatKit now runs on consumer GPUs with a new 7B parameter model

Loving this gold rush for self hosted ML models!