Mistral-Large-Instruct-2407 Released: Multilingual AI with 128K Context, 80+ Coding Languages, 84.0% MMLU, 92% HumanEval, and 93% GSM8K Performance

We are a community of AI/ ML/Generative AI enthusiasts/researchers/journalists/writers who share interesting news and articles about the applications of AI. 

Machine Learning News

Mistral AI has released its new flagship model, Mistral Large 2, which exhibits notable advancements in code generation, mathematics, and multilingual support. Featuring a 128K context window and supporting over 80 coding languages, the model is optimized for single-node inference with 123 billion parameters. It achieves high accuracy on benchmarks like MMLU and HumanEval and is available for research and commercial use under respective licenses. The model can be accessed on 'la Plateforme' and is integrated with major cloud providers including Google Cloud Platform, Azure AI Studio, Amazon Bedrock, and IBM watsonx.ai.