Turkish Llama 8b Instruct V0.1 GGUF

🚀 Turkish-Llama-8b-Instruct-v0.1 Quantized Model

This project focuses on the quantization of the Turkish-Llama-8b-Instruct-v0.1 model and provides a performance analysis of Turkish language support models.

🚀 Quick Start

This README provides detailed information about the Turkish-Llama-8b-Instruct-v0.1 quantized model, including base model information, license details, performance analysis of Turkish language support models, and download links for quantized models. It also lists various applications and tools for locally quantized LLMs.

✨ Features

• Base Model: ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1
• License: llama3
• Quantization: Quantized with llama.cpp using all-gguf-same-where
• Performance Analysis: Comprehensive testing of multiple Turkish language support models based on specific criteria.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

No code examples are provided in the original document.

📚 Documentation

🔍 Tested Models

The following local models were comprehensively tested for their Turkish text generation and understanding capabilities:

1. CohereForAI/c4ai-command-r7b-arabic-02-2025
2. CohereForAI/c4ai-command-r7b-12-2024
3. CohereForAI/aya-expanse-8b
4. ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1
5. google/gemma-3-4b-it
6. google/gemma-3-4b-pt-qat-q4_0-gguf

🔍 Test Criteria

The models were evaluated using the following 3 basic questions:

1. "Can you explain the 3 most important historical artifacts in Turkey and their cultural significance?"
2. "Compare the differences between the Asian and European sides of Istanbul in English/Turkish"
3. "Create a 3-day tourist plan for Antalya"

📊 Outstanding Results

🔗 Full Test Results

To examine all model responses in detail:
📂 Turkish-Supported Model Comparisons

💡 User Guide

- 🔥 **Turkish-Priority Projects** → `Turkish-Llama-8B` (Highest cultural fit)
- 🌐 **Multilingual Applications** → `Aya-Expanse-8B`  
- 📜 **Official Documents** → `Command R7B` series
- 📱 **Low-Resource** → `Gemma-3-4B` (Compatible with Android)

⚠️ Important Note

• All models were tested in the GGUF quantized format.
• There may be a 5-10% error margin in geographical names.
• Ollama/LM Studio is recommended for local use.

🌟 Personal Experience

"During my tests with local models, especially Turkish-Llama-8B impressed me with its natural use of Turkish and mastery of cultural context. You can also try these models in your own projects and share the results on our AnkaNLP page!"

✅ Quantized Models Download List

🔍 Recommended Quantizations

• ✨ General CPU Use: Q4_K_M (Best balance of speed/quality)
• 📱 ARM Devices: Q4_0 (Optimized for ARM CPUs)
• 🏆 Maximum Quality: Q8_0 (Near-original quality)

📦 Full Quantization Options

🚀 Download	🔢 Type	📝 Notes
Download		Basic quantization
Download		Small size
Download		Balanced quality
Download		Better quality
Download		Fast on ARM
Download		Fast, recommended
Download	⭐	Best balance
Download		Good quality
Download		Balanced
Download		High quality
Download	🏆	Very good quality
Download	⚡	Fast, best quality
Download		Maximum accuracy

💡 Usage Tip

Use F16 for maximum precision when quality is critical

🚀 Applications and Tools for Locally Quantized LLMs

🖥️ Desktop Applications

📱 Mobile Applications

🎨 Image Generation Applications

📄 License

The project is licensed under the llama3 license.