“Zero One Everything 01AI" official WeChat account announced tonightOpen Source Yi-9B Model, officially called the Yi seriesModelThe "Science Champion" in the series - Yi-9B is the model with the strongest code and mathematical capabilities in the current Yi series models.RealityThe actual parameter is 8.8B, and the default context length is 4K tokens..
This model is based on Yi-6B (trained with 3.1T tokens) and uses 0.8T tokens for further training. The data is as of June 2023.
According to reports, in terms of comprehensive ability (Mean-All),Yi-9B's performance is "better than other similar sizedOpen SourceThe best model, surpassing DeepSeek-Coder, DeepSeek-Math, Mistral-7B, SOLAR-10.7B and Gemma-7B.
In terms of coding capability (Mean-Code), Yi-9B’s performance is second only to DeepSeek-Coder-7B, and surpasses Yi-34B, SOLAR-10.7B, Mistral-7B, and Gemma-7B.
In terms of mathematical capabilities (Mean-Math), Yi-9B's performance is second only to DeepSeek-Math-7B, and surpasses SOLAR-10.7B, Mistral-7B, and Gemma-7B.
In terms of common sense and reasoning ability (Mean-Text), Yi-9B performs on par with Mistral-7B, SOLAR-10.7B, and Gemma-7B.
The official said,Yi-9B (BF 16) and its quantized version Yi-9B (Int8) can be easily deployed on consumer-grade graphics cards, low cost of use and developer-friendly.
The companyKai-Fu LeeThe chairman and CEO of Sinovation Ventures led the team to found the company, which has already launched two models, Yi-34B and Yi-6B.Open Source Big Model, claiming to be completely open to academic research and simultaneously open to free commercial applications.