---
language:
- en
license: apache-2.0
base_model: mistralai/Mistral-7B-Instruct-v0.2

datasets:
- Antonio88/TaliStran-DataSet
tags:
- Conversational
---


![image/png](https://cdn-uploads.huggingface.co/production/uploads/644fc68af8b353c9491785d8/ARiB2V3Zgj1b7b26lvnWM.png)



TaliStran-7B-beta-V.1: The Italian Language Model

TaliStran is a model developed by leveraging the power of Mistral 7B Instruct v0.2. This model has been specifically trained to understand and interact in Italian, providing coherent responses to questions asked in this language.

Main Features:
Training: TaliStran was trained on a selected small dataset of about 500 questions and answers, ensuring targeted preparation for understanding and generating text in Italian.

Training Efficiency: The training process was carried out using Google Colab, utilizing an A100 graphics card, which allowed the preparation to be completed efficiently and with limited resources.

Model in Test

there will be continuous improvements