Llama3 1 Turkish ChatBot

🚀 Llama3.1 Turkish Educational ChatBot

This project presents a fine - tuned version of the Meta LLaMA 3.1 8B large language model. It's specifically tailored to handle Turkish educational question - answer prompts. The core objective is to offer fluent, informative, and context - aware Turkish answers, which are ideal for general inquiries and learning support.

The model was fine - tuned using the LoRA technique on a small scale (1% of trainable parameters) and is available on Hugging Face:

👉 [metehanayhan/Llama3 - 1_Turkish_ChatBot](https://huggingface.co/metehanayhan/Llama3 - 1_Turkish_ChatBot)

🚀 Quick Start

from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

tokenizer = AutoTokenizer.from_pretrained("metehanayhan/Llama3-1_Turkish_ChatBot")
model = AutoModelForCausalLM.from_pretrained("metehanayhan/Llama3-1_Turkish_ChatBot")

qa_pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
print(qa_pipe("İklim değişikliği neden önemlidir?", max_new_tokens=200)[0]["generated_text"])

✨ Features

• This is a fine - tuned version of Meta LLaMA 3.1 8B, specialized for Turkish educational Q&A.
• It can provide fluent, informative, and context - aware answers in Turkish.
• The model uses the LoRA technique for fine - tuning, enabling effective learning with low computational resources.

📦 Installation

The installation details are not provided in the original README. However, based on the quick - start code, you need to have adapter - transformers installed. You can install it using pip install adapter - transformers.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

tokenizer = AutoTokenizer.from_pretrained("metehanayhan/Llama3-1_Turkish_ChatBot")
model = AutoModelForCausalLM.from_pretrained("metehanayhan/Llama3-1_Turkish_ChatBot")

qa_pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
print(qa_pipe("İklim değişikliği neden önemlidir?", max_new_tokens=200)[0]["generated_text"])

Advanced Usage

There is no advanced usage example in the original README. You can potentially adjust parameters such as max_new_tokens and other hyperparameters according to your specific needs.

📚 Documentation

Training Summary

Data Format

The dataset consists of Turkish question - answer pairs in CSV format. Each row represents a complete educational sample.

Example:

question,answer
What can be done to prevent climate change?,
"To combat climate change, actions like reducing fossil fuel usage and transitioning to renewable energy sources are essential..."

A total of 17,587 such examples were used for fine - tuning.

Use Case

This model serves as an educational assistant in Turkish. It can answer questions related to:

• Informative, general - knowledge, or school - related topics
• Support for curious learners and students

Deployment

The model is shared on Hugging Face with 4 - bit quantization and is ready for low - resource inference. It has also been exported in GGUF format for use in compatible environments.

Additional Notes

• The training was performed using Trainer with standard SFT configuration.
• random_state = 3407 was used to ensure reproducibility.
• Although fine - tuned on just 1% of parameters, the model responds effectively across a wide range of Turkish topics.

🔧 Technical Details

• The model is a fine - tuned version of Meta LLaMA 3.1 8B.
• The LoRA technique is used for fine - tuning, which only trains 1% of the model's parameters.
• The training data is Turkish question - answer pairs in CSV format.
• The model is quantized to 4 - bit for low - resource inference.

📄 License

The model is released under the MIT license.