 | Interestingly enough, Mistral Small is written as Mistral-Small-4-119B-2603. Their medium model will have 128B paramters. Either it will be a dense model, or a less sparse MoE than Mistral Small [link] [comments] |
| | Interestingly enough, Mistral Small is written as Mistral-Small-4-119B-2603. Their medium model will have 128B paramters. Either it will be a dense model, or a less sparse MoE than Mistral Small [link] [comments] |
![[link]](https://i.redd.it/fxlkj3f6xyxg1.png){kind=link}