KEMU-GPT

KEMU-GPT is an advanced language model based on GPT-2 Medium and focused entirely on the English language. It is an AI without any subject limitations; it is designed to answer any question you can think of

Model Details

• Base Model: GPT-2 Medium
• Parameters: ~355M
• Training: LoRA fine-tuning
• Languages: English
• Size: ~1.5GB

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mustafakemal0146/kemu-gpt")
model = AutoModelForCausalLM.from_pretrained("mustafakemal0146/kemu-gpt")

# Generate text
prompt = "Merhaba, nasılsın?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=100)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Training Data

The model was fine-tuned on a curated dataset focusing on:

• English philosophy books
• General knowledge questions
• Helpful assistant responses

Limitations

• This model is for educational and research purposes
• May generate biased or inappropriate content
• Users should verify important information

License

MIT License - see LICENSE file for details.

Citation

@misc{kemu-gpt,
  title={KEMU-GPT: A Fine-tuned English Language Model},
  author={Mustafa Kemal Çıngıl},
  year={2025},
  url={https://huggingface.co/mustafakemal0146/kemu-gpt}
}