# Mistral fine-tuning

This model is an experiment in the field of formal theorem proving, specifically designed for generating and explaining Coq code. Utilizing a comprehensive dataset derived from over 10,000 Coq source files, it demonstrates enhanced capabilities in understanding Coq-specific syntax and semantics, thereby driving significant advancements in automated theorem proving.

Large open-source Dutch conversational model based on Mistral 7B, significantly improving Dutch language capabilities through additional training on 10 billion Dutch text tokens

PiVoT is a model fine-tuned from Mistral 7B, derived from Synatra v0.3 RP variant, demonstrating good performance.