view article Article Fine-Tune ViT for Image Classification with π€ Transformers Feb 11, 2022 β’ 40
view post Post 1479 You are happy that @mistralai is releasing a new model πYou become even more happy to see it's a completely new coding model πThen you become sad because the model is licensed under MNLP πBefore we talk about MNLP, here is the gist of the model:π·οΈName: Codestral (Code + Mistral π)π 22B parametersπ Supports 80 programming languages (including Python, Java, C, C++, bash, swift, and more)π Outperforms Llama 3 70B and Code Llama 70B on HumanEval and MBPPπ Outperforms DeepSeek Coder 33B on HumanEvalπ 32K context window (longer than Llama 3, DeepSeek, or Code Llama)π€ Supports both code assistant and code completion use casesMore details: https://mistral.ai/news/codestral/ mistralai/Codestral-22B-v0.1Now what's MNLP? It's a non-commercial license for Mistral models that Codestral is released under! More here: https://mistral.ai/news/mistral-ai-non-production-license-mnpl/Don't be sad... π There is another model that's open source and actually gives better performance on HumanEval: Bin12345/AutoCoder π 5 5 π 2 2 + Reply