A new variant of the Mistral AI model, named 'Mistral 7B Fine-Tune Optimized,' has been released, boasting superior performance compared to other 7B models and even surpassing GPT-4 from March in evaluations. The model has been praised for its ability to be a strong base for downstream fine-tuning. According to user reports, fine-tunes trained on this new variant are stronger than those trained on any other 7B model. In addition to its performance, the model has been recognized for its utility and state-of-the-art status in nearly every category. The developers have updated their website to provide more information and facilitate community engagement. Furthermore, the new Mistral AI model, Mixtral 8x7B, has shown impressive SQL capabilities when running on platforms provided by TogetherCompute x LangChainAI. User experiences indicate that the model is highly responsive to a wide range of questions, including unsafe ones, and can provide answers in various summary lengths. It is noted as a combination of two other Mistral 7B models and has achieved a 46% score on humaneval.
I got around to trying this model and all I can say is... you can get it to answer *any* question, even very unsafe ones. It follows instructions surprisingly well (ex. give me a short, medium & long summary). It's a merge of two other mistral-7B models, also got 46% on humaneval https://t.co/zV18clKQJC https://t.co/etzs5JFW8u
https://t.co/1lHmACFf6u https://t.co/Asb4L2haXr in case you guys missed the previous announcement! a 7b that outperforms gpt4 (from march) at humaneval!, a ton of utility! Check out the new https://t.co/hqVW3tD61o ! the website so you guys can more easily follow up, join… https://t.co/dadGetoM8t
https://t.co/1lHmACFf6u https://t.co/Asb4L2haXr in case you guys missed the previous announcement! a 7b that outperforms gpt4 (from march) at humaneval!, a ton of utility! not to mention state of the art in near every category to boot! we've updated the website so you… https://t.co/5y6JpO3elw
https://t.co/1lHmACFf6u https://t.co/Asb4L2haXr in case you guys missed the previous announcement! a 7b that outperforms gpt4 (from december) at humaneval!, a ton of utility! not to mention state of the art in near every category to boot! we've updated the website so you… https://t.co/KcLNQXbxUE
The new @MistralAI Mixtral 8x7B is doing a great SQL. Running on @togethercompute x @LangChainAI Source:- https://t.co/O4namwDpG1 https://t.co/c0606YTHHC
Just dropped a new Mistral variant "Mistral 7B Fine-Tune Optimized." Built to be a very strong base for downstream fine-tuning. On our evals fine-tunes trained on it are stronger than fine-tunes on any other 7B model, and even beat gpt-4. 🙂 https://t.co/SyhqFAJz5y