Instructions to use Edraky/Edraky-AI with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Edraky/Edraky-AI with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Edraky/Edraky-AI")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Edraky/Edraky-AI")
model = AutoModelForCausalLM.from_pretrained("Edraky/Edraky-AI")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Edraky/Edraky-AI with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Edraky/Edraky-AI"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Edraky/Edraky-AI",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Edraky/Edraky-AI

SGLang

How to use Edraky/Edraky-AI with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Edraky/Edraky-AI" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Edraky/Edraky-AI",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Edraky/Edraky-AI" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Edraky/Edraky-AI",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Edraky/Edraky-AI with Docker Model Runner:
```
docker model run hf.co/Edraky/Edraky-AI
```

Edraky commited on Jul 21, 2025

Commit

721eda9

verified ·

1 Parent(s): 025311d

Create Edraky AI!

Browse files

<p align="center">
<img src="https://cdn-uploads.huggingface.co/production/uploads/686e726239f003427404a1be/uuB7LFKDX1C5B28DGJyZN.png" alt="Edraky Logo" width="200"/>
</p>

# 🧠 إدراكي (Edraky) — Multilingual Educational AI Assistant

**إدراكي (Edraky)** is a multilingual AI model designed to support students with interactive learning, question-answering, and intelligent content generation across **Arabic**, **English**, and **Hebrew**. It is especially focused on **educational content** for middle-school learners, including subjects like Social Studies, Arabic Language, and more.

---

## 🧾 Model Details

- **Languages**: Arabic, English, Hebrew
- **License**: Apache 2.0
- **Base Model**: `Qwen/Qwen2-1.5B-Instruct`
- **Library**: `transformers`
- **Pipeline Tag**: `text-generation`
- **Trained on**:
- `rosoft/rStar-Coder`
- `gsm8k-rerun/Qwen_Qwen2.5-1.5B-Instruct`
- **Tags**: `multilingual`, `educational`, `fine-tuned`, `arabic`, `hebrew`, `qwen`, `student-assistant`

---

## 🎯 Use Cases

- 📚 **AI Study Assistant**: Helps students with curriculum-based questions
- 🧠 **Educational Content Creation**: Generates quiz questions, summaries, and explanations
- 🌐 **Language Support**: Understands and responds in Arabic, English, and Hebrew
- 👨‍🏫 **Teacher Support**: Generates model answers, revision notes, and more

---

## 🔍 Evaluation Metrics

| Metric | Value |
|------------|-------|
| Accuracy | ✅ High (on educational QA datasets) |
| BLEU Score | 🧪 Coming Soon |
| F1 Score | 🧪 Coming Soon |

---

## 📚 Training Datasets

The model was fine-tuned on high-quality datasets including:

- [`rosoft/rStar-Coder`](https://huggingface.co/datasets/rosoft/rStar-Coder)
- [`gsm8k-rerun/Qwen_Qwen2.5-1.5B-Instruct`](https://huggingface.co/datasets/gsm8k-rerun/Qwen_Qwen2.5-1.5B-Instruct)

Additional in-house educational data tailored for Egyptian curricula was used to optimize performance.

---

## 🚀 How to Use

You can try it with the Hugging Face Inference API or use it directly via Transformers:

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "your-username/edraky-ai"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

inputs = tokenizer("اشرح درس الثورة العرابية", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
💡 Intended Uses
This model is designed to be:

🧑‍🎓 Student-friendly

👩‍🏫 Useful for teachers and educators

🌐 Ready for multilingual educational tasks

⚠️ Limitations
This model is not a replacement for a certified teacher.

Knowledge is limited to the data it was trained on.

May hallucinate or generate incorrect facts; always verify educational material.

🤝 Contributing
We're open to collaboration, dataset contributions, or feedback to improve this model further for educational use in the Arab world and beyond.

🧾 License
Apache License 2.0 — Free to use, share, and modify with attribution.

Files changed (1) hide show

README.md +26 -3

README.md CHANGED Viewed

@@ -1,3 +1,26 @@
----
-license: apache-2.0
----

+---
+license: apache-2.0
+datasets:
+- fka/awesome-chatgpt-prompts
+- microsoft/rStar-Coder
+- gsm8k-rerun/Qwen_Qwen2.5-1.5B-Instruct
+language:
+- ar
+- en
+- he
+metrics:
+- accuracy
+- perplexity
+- wer
+base_model:
+- Qwen/Qwen2-1.5B-Instruct
+pipeline_tag: text-generation
+library_name: transformers
+tags:
+- multilingual
+- arabic
+- hebrew
+- qwen
+- educational
+- fine-tuned
+---