description Yi-68B Overview
Yi-68B is a powerful open-source LLM developed by 01.AI. It stands out for its exceptionally long context window (up to 32K tokens), enabling it to process and generate text based on significantly larger inputs than many competitors. This makes it ideal for tasks requiring deep understanding of complex documents or conversations.
help Yi-68B FAQ
Who created the Yi-68B language model?
Yi-68B was released by 01.AI, the company founded by AI researcher Kai-Fu Lee. It belongs to the original Yi family alongside the smaller Yi-6B model.
How much hardware is needed to run Yi-68B locally?
Running all 68 billion parameters at 16-bit precision requires well over 130 GB of memory before runtime overhead. A 4-bit quantization is much smaller but still typically needs roughly 40 GB or more across GPU memory and system RAM.
What context length does Yi-68B support?
The original Yi-68B release supports a context window of up to 4K tokens, while the separately released Yi-68B-200K extends that to 200K. A catalog claim of a universal 32K limit therefore does not identify the model variants correctly.
Is Yi-68B a chat model or a base model?
Both forms were released: Yi-68B is the pretrained base model, and Yi-68B-Chat is tuned for conversational instruction following. Self-hosters should choose the chat variant for an assistant interface and the base model for further training or completion workflows.
explore Explore More
Similar to Yi-68B
ui.x_see_all arrow_forward
Reviews & Comments
Write a Review
Be the first to review
Share your thoughts with the community and help others make better decisions.
