Model Overview

malhajar/Mistral-7B-Instruct-v0.2-turkish is a 7 billion parameter instruction-tuned language model developed by Mohamad Alhajar. It is a specialized adaptation of the Mistral-7B-Instruct-v0.2 base model, finetuned using SFT Training and the Freeze method.

Key Capabilities

• Turkish Language Proficiency: Specifically optimized for understanding and generating responses in Turkish.
• Instruction Following: Designed to accurately follow and respond to instructions in a chat-based format.
• Chatbot Applications: Well-suited for conversational AI tasks where Turkish language interaction is required.

Training Details

The model was finetuned on the alpaca-gpt4-tr dataset, which focuses on Turkish instructions. This targeted training enhances its performance for Turkish-specific use cases, differentiating it from its base model.

Prompt Format

Users should structure prompts using the following template for optimal performance:

### Instruction:
<prompt>
### Response:

This model is ideal for developers looking to integrate a capable, Turkish-centric instruction-following LLM into their applications.