TurkishReasoner-Llama3.1-8B

TurkishReasoner-Llama3.1-8B

Model Description

TurkishReasoner-Llama8B leverages Meta's powerful Llama3.1-8B foundation model to deliver sophisticated reasoning capabilities in Turkish. Fine-tuned using GRPO techniques, this model excels at multistep reasoning processes with particular strength in mathematical problem-solving and logical deduction.

Key Features

• Built on Meta's advanced Llama3.1-8B foundation
• Optimized for Turkish reasoning tasks with structured output
• Balanced performance-to-resource ratio (8B parameters)
• Strong multilingual understanding with Turkish specialization
• Trained using Group Relative Policy Optimization (GRPO)
• Clear step-by-step reasoning with formatted solutions

Technical Specifications

• Base Model: Meta/Llama3.1-8B
• Parameters: 8 billion
• Input: Text
• Hardware Requirements: ~16GB VRAM
• Training Infrastructure: NVIDIA Ada6000 GPU

Usage

This model is well-suited for a variety of Turkish reasoning applications:

• Educational platforms requiring detailed explanations
• Research tools analyzing complex problem-solving approaches
• Development of Turkish-language assistants with robust reasoning
• Applications requiring balanced performance and efficiency

Example Usage

from transformers import pipeline

pipe = pipeline("text-generation", model="Chan-Y/TurkishReasoner-Llama3.1-8B", device=0)

messages = [
    {"role": "system", "content": """Sen kullanıcıların isteklerine Türkçe cevap veren bir asistansın ve sana bir problem verildi.
Problem hakkında düşün ve çalışmanı göster.
Çalışmanı <start_working_out> ve <end_working_out> arasına yerleştir.
Sonra, çözümünü <SOLUTION> ve </SOLUTION> arasına yerleştir.
Lütfen SADECE Türkçe kullan."""},
    {"role": "user", "content": "121'in karekökü kaçtır?"},
]

response = pipe(messages)
print(response)

For more information or assistance with this model, please contact the developers:

• Cihan Yalçın: https://www.linkedin.com/in/chanyalcin/
• Şevval Nur Savcı: https://www.linkedin.com/in/%C5%9Fevval-nur-savc%C4%B1/