---
license: cc-by-nc-nd-4.0
language:
- en
- tr
metrics:
- accuracy
pipeline_tag: question-answering
library_name: diffusers
---

📌 General Information:
🖥 Programming Language: PYTHON (Py/py/PY)
📂 Model Code File: Br-T-1-m-pre.py (Located in the "Files and versions" section.)
📚 Dataset: "wikitext-2-raw-v1"
📏 Context Length: 64 characters / 48 tokens
🔢 Number of Parameters: 15 million
🎯 Number of Attention Heads: 8
💡 Use Cases:
✍ Sentence Completion
❓ Answer Prediction
🤖 QA (Question Answering)
🔥 Performance Improvement:
Br-T-1-m-pre achieves 13x higher MMLU scores compared to previous versions!

💲 Pricing:
💰 $0 / 0₺ / 0£ / 0€ → The model is COMPLETELY FREE!

🚫 Restrictions:
The model must not be used for:
🔴 Sexual, violent, suicide-related, obscene, racist, drug-related, self-harm, smuggling, terrorism, copyright-infringing, digital attack content
🔴 Any unethical and immoral requests

⚠ Commercial Usage:
❌ Commercial use is prohibited!

📢 Attribution:
✅ You must provide attribution! (Br-T-1-m-pre by Bertuğ Günel)

👨‍💻 Development Team:
Lead Developer: Bertuğ Günel
Location: Turkey / Eskişehir

![Figure_1.png](https://cdn-uploads.huggingface.co/production/uploads/66eeed95df079df1a35a553b/-XRNk46FJu9njA34RVoNA.png)