The community member was looking at the MTBE benchmark and saw that an LLM called Mistral, with 7B parameters, was ranked number one. They asked if they could use the fine-tuning repository to fine-tune Mistral. In the comments, another community member advised against using an LLM as an embedding model, stating that there is currently no support for fine-tuning LLMs for this purpose. Another comment noted that the 7B parameter Mistral model only barely beats other models.
