Yehia: A Simple (nice to talk to) Arabic Model
๐ค What is Yehia?
Yehia is a 7-billion-parameter language model built to be more than just a toolโitโs a companion. Based on ALLaM-AIโs ALLaM-7B-Instruct-preview, Yehia is designed to offer thoughtful, kind, and helpful conversations in both Arabic and English.
You can chat with Yehia from here ๐
๐ฐ Interesting News
As of 2/3/2025, Yehia is the best Arabic model on AraGen-Leaderboard between the sizes of 0.5B up to 25B ๐ฅ
๐ ๏ธ How Yehia was made?
Yehia is trained using Group Relative Policy Optimization (GRPO) โa method that refines its answers by comparing and selecting the best responses. Its development follows the 3C3H metric, prioritizing:
- Correctness โ : Accurate information to build trust.
- Completeness ๐: Full, well-rounded answers.
- Conciseness โ๏ธ: Clear, to-the-point responses.
- Helpfulness ๐ค: Always aiming to support and uplift.
- Honesty ๐ฌ: Transparent, straightforward communication.
- Harmlessness โค๏ธ: Promoting kindness and safety.
And the Judge model of our answer was none other than claude-sonnet-3.5 ๐
๐ Getting Started
To start using Yehia, you can easily load the model with the transformers library:
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
model_name = "Navid-AI/Yehia-7B-preview"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, attn_implementation="flash_attention_2", device_map="auto")
messages = [
{"role": "system", "content": "ุฃูุช ูุญููุ ุฐูุงุกู ุงุตุทูุงุนููู ุทูุฑุชู ุดุฑูุฉ 'ูููุฏ'ุ ู
ุชุฎุตุตู ูู ุงูุชูููุฑ ุงูู
ูุทูู ูุงูุชุญููู ุงูุฏููู. ู
ูู
ุชู ุฅููุงู
ุงูู
ุณุชุฎุฏู
ูู ูุฏุนู
ูู
ูู ุฑุญูุชูู
ูุญู ุงูุชุนููู
ุ ุงููู
ูุ ูุชุญููู ุฃูุฏุงููู
ู
ู ุฎูุงู ุชูุฏูู
ุญูููู ุฐููุฉู ูู
ุฏุฑูุณุฉ."},
{"role": "user", "content": "ู
ุฑุญุจุงู ูุง ูุญูู! ููู ุญุงูู ุงูููู
ุ"}
]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt", return_dict=True).to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
Note: If flash_attention_2 is giving you any problems just remove it.
๐ What Can Yehia Do?
- Explain Concepts ๐ก: Break down educational topics in Arabic to help learners understand easily.
- Engage in Conversations ๐ฃ๏ธ: Offer friendly and supportive chats that uplift users.
- Promote Learning ๐: Encourage curiosity and provide knowledge in an accessible way.
Yehia shines in conversations that feel personal and uplifting, always striving to improve.
๐ญ Remember
Yehiaโs name means โGod is graciousโ in Arabicโreflecting its mission to bring grace and connection to every interaction. Whether youโre a student, creator, or just curious, Yehia is here to brighten your day.
๐ Citation
If you would like to cite Yehia in your work, please use the following BibTeX entry:
@misc{yehia2025,
title={Yehia 7B Preview},
author={Navid-AI},
year={2025},
howpublished={\url{https://huggingface.co/Navid-AI/Yehia-7B-preview}}
}
- Downloads last month
- 388