Turkish-LLM-7B-Instruct

The first open-source instruction-tuned Turkish language model at 7B scale.

Highlights

Native Turkish - Trained specifically for Turkish language tasks
Instruction Following - Optimized for chat and Q&A
7B Parameters - Balanced performance and efficiency
Open Source - Apache 2.0 License

Model Details

Attribute	Value
Developer	Ogulcan Aydogan
Base Model	TURKCELL/Turkcell-LLM-7b-v1
Parameters	7 Billion
Language	Turkish (tr)
License	Apache 2.0
Fine-tuning	LoRA (Low-Rank Adaptation)
Training Data	125,000+ Turkish instruction-response pairs

Model Family

Model	Parameters	Base	Method	Use Case
Turkish-LLM-14B-Instruct	14.7B	Qwen2.5-14B-Instruct	SFT	Higher quality, complex reasoning
Turkish-LLM-14B-Instruct-GGUF	14.7B	14B-Instruct	GGUF quantized	Local/edge deployment
Turkish-LLM-7B-Instruct (this)	7B	Turkcell-LLM-7b-v1	LoRA	Lightweight, faster inference

Training

Parameter	Value
Hardware	NVIDIA A100 80GB
Framework	PyTorch + Transformers + PEFT
Precision	bfloat16
Final Loss	1.88
Learning Rate	5e-6
Batch Size	16
Max Sequence Length	2048
LoRA Rank	64
LoRA Alpha	128

Usage

Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained(
    "ogulcanaydogan/Turkish-LLM-7B-Instruct",
    torch_dtype=torch.bfloat16,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("ogulcanaydogan/Turkish-LLM-7B-Instruct")

messages = [
    {"role": "user", "content": "Turkiye'nin baskenti neresidir?"}
]

prompt = "<|im_start|>user\n" + messages[0]["content"] + "<|im_end|>\n<|im_start|>assistant\n"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(
    **inputs,
    max_new_tokens=512,
    temperature=0.7,
    top_p=0.9,
    do_sample=True
)
print(tokenizer.decode(outputs[0], skip_special_tokens=True).split("assistant\n")[-1])

Ollama

ollama run hf.co/ogulcanaydogan/Turkish-LLM-7B-Instruct

Chat Template

<|im_start|>user
{user_message}<|im_end|>
<|im_start|>assistant
{assistant_response}<|im_end|>

Example Outputs

Q: Turkiye'nin baskenti neresidir? A: Turkiye'nin baskenti Ankara'dir.

Q: Yapay zeka nedir? A: Yapay zeka, ogrenme ve akil yurutme yetenegine sahip bilgisayar sistemlerini ifade eder.

Hardware Requirements

Precision	VRAM Required	Recommended
BF16	~14 GB	RTX 4090, A10G, M2 Pro (16GB)
INT8	~7 GB	RTX 3080, M1 Pro
INT4	~4 GB	RTX 3060, Apple M-series (8GB)

Intended Use

Turkish chatbots and virtual assistants
Question answering systems
Text generation and creative writing
Educational applications
NLP research for Turkish language

Limitations

May occasionally generate incorrect information (hallucinations)
Performance on very long contexts (>2048 tokens) may degrade
Not recommended for production without additional safety measures

Related Resources

Resource	Link
14B Model	Turkish-LLM-14B-Instruct
14B GGUF	Turkish-LLM-14B-Instruct-GGUF
Live Demo (14B)	Turkish-LLM-14B-Chat
Live Demo (7B)	Turkish-LLM-7B-Chat
Training Pipeline	LowResource-LLM-Forge
Project Repository	Turkish-LLM on GitHub

Citation

@misc{turkish-llm-7b-instruct-2026,
  author = {Aydogan, Ogulcan},
  title = {Turkish-LLM-7B-Instruct: An Instruction-Tuned Turkish Language Model},
  year = {2026},
  publisher = {HuggingFace},
  url = {https://huggingface.co/ogulcanaydogan/Turkish-LLM-7B-Instruct}
}