File size: 6,367 Bytes
c601505 a0126ed 40412e8 a0126ed 40412e8 a0126ed 40412e8 a0126ed |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 |
---
license: cc-by-nc-4.0
datasets:
- hammh0a/Hala-4.6M-SFT
language:
- ar
base_model:
- QCRI/Fanar-1-9B-Instruct
pipeline_tag: text-generation
---
# Hala: Arabic‑Centric Instruction & Translation Models
<p align="center">
<img src="https://i.ibb.co/pvhp1XfJ/halalogo.png" alt="Hala logo" width="550" />
</p>
**Paper**: *Hala Technical Report: Building Arabic‑Centric Instruction & Translation Models at Scale*
**Authors**: Hasan Abed Al Kader Hammoud\*, Mohammad Zbeeb\*, Bernard Ghanem
**Affiliation**: King Abdullah University of Science and Technology (KAUST)
\*Equal contribution
> In Arabic, **حلا** (Hala) conveys sweetness and beauty—qualities long associated with the language itself. In this spirit, we call our models **Hala**.
---
## 🔗 Quick Links
* **Models & Data (Hugging Face collection)**: [https://huggingface.co/collections/hammh0a/hala-68bf02b34a14b9f22305ab3a](https://huggingface.co/collections/hammh0a/hala-68bf02b34a14b9f22305ab3a)
* **Contact**: [hasanabedalkader.hammoud@kaust.edu.sa](mailto:hasanabedalkader.hammoud@kaust.edu.sa)
---
## Example
```python
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
model_id = "hammh0a/Hala-9B" # pick a released Hala model
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id, torch_dtype="auto", device_map="auto"
)
# Use chat template
messages = [
{"role": "system", "content": "أنت مساعد خبير في الفيزياء."},
{"role": "user", "content": "اشرح بإيجاز مبدأ الانحفاظ في الفيزياء، وأعطني مثالاً يومياً."},
]
prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipe = pipeline("text-generation", model=model, tokenizer=tok)
out = pipe(prompt, max_new_tokens=256, do_sample=False)
print(out[0]["generated_text"])
```
---
## 📊 Results
*Hala models are placed at the end of each size category; best **Average** per category is in bold.*
### ≤2B parameters
| Size | Model Name | Params | AlGhafa | ArabicMMLU | EXAMS | MadinahQA | AraTrust | ArbMMLU‑HT | Average |
| ---- | -------------------------------------- | -----: | ------: | ---------: | ----: | --------: | -------: | ---------: | -------: |
| ≤2B | meta-llama/Llama-3.2-1B | 1B | 33.9 | 26.5 | 21.2 | 25.7 | 37.1 | 23.9 | 28.0 |
| ≤2B | Qwen/Qwen2-1.5B-Instruct | 1.5B | 53.1 | 49.2 | 35.2 | 45.5 | 68.9 | 37.4 | 48.2 |
| ≤2B | Qwen/Qwen2.5-1.5B-Instruct | 1.5B | 48.4 | 43.5 | 31.8 | 38.2 | 70.8 | 35.9 | 44.8 |
| ≤2B | Sakalti/Saka-1.5B | 1.5B | 51.4 | 40.0 | 31.3 | 31.5 | 47.5 | 33.5 | 39.2 |
| ≤2B | Qwen/Qwen3-1.7B-Base | 1.7B | 56.8 | 49.7 | 38.2 | 40.0 | 75.6 | 43.9 | 50.7 |
| ≤2B | Qwen/Qwen1.5-1.8B | 1.8B | 32.7 | 26.7 | 23.8 | 26.0 | 31.5 | 23.6 | 27.4 |
| ≤2B | silma-ai/SILMA-Kashif-2B-Instruct-v1.0 | 2B | 59.7 | 45.6 | 33.1 | 38.8 | 73.3 | 35.8 | 47.7 |
| ≤2B | google/gemma-2-2b-it | 2B | 34.1 | 30.1 | 23.6 | 20.1 | 31.2 | 23.4 | 27.1 |
| ≤2B | LiquidAI/LFM2-350M | 350M | 39.0 | 35.2 | 30.9 | 28.3 | 43.3 | 29.1 | 34.3 |
| ≤2B | **Hala‑350M** | 350M | 51.4 | 41.2 | 36.9 | 34.5 | 52.1 | 35.4 | 41.9 |
| ≤2B | LiquidAI/LFM2-700M | 700M | 50.1 | 38.3 | 34.3 | 32.5 | 56.3 | 37.2 | 41.4 |
| ≤2B | **Hala‑700M** | 700M | 55.5 | 45.9 | 40.6 | 34.7 | 65.2 | 39.4 | 46.9 |
| ≤2B | LiquidAI/LFM2-1.2B | 1.2B | 53.8 | 45.2 | 35.0 | 34.7 | 65.6 | 43.4 | 46.3 |
| ≤2B | **Hala‑1.2B** | 1.2B | 59.2 | 48.6 | 43.4 | 41.6 | 71.7 | 44.2 | **51.4** |
### 7B–9B parameters
| Size | Model Name | Params | AlGhafa | ArabicMMLU | EXAMS | MadinahQA | AraTrust | ArbMMLU‑HT | Average |
| ----- | ------------------------------------------- | -----: | ------: | ---------: | ----: | --------: | -------: | ---------: | -------: |
| 7B–9B | CohereForAI/c4ai-command-r7b-arabic-02-2025 | 7B | 74.8 | 59.3 | 65.0 | 63.8 | 80.5 | 50.1 | 65.6 |
| 7B–9B | JasperV13/Yehia-7B-DPO-Reasoning-preview | 7B | 75.1 | 66.3 | 51.8 | 54.9 | 81.9 | 55.1 | 64.2 |
| 7B–9B | Navid-AI/Yehia-7B-preview | 7B | 70.8 | 64.9 | 52.1 | 54.4 | 87.5 | 53.4 | 63.9 |
| 7B–9B | JasperV13/Yehia-7B-Reasoning-preview | 7B | 75.2 | 66.3 | 52.7 | 55.0 | 80.8 | 55.2 | 64.2 |
| 7B–9B | ALLaM-AI/ALLaM-7B-Instruct-preview | 7B | 69.5 | 64.9 | 51.6 | 54.2 | 86.9 | 52.8 | 63.3 |
| 7B–9B | Qwen/Qwen2-7B-Instruct | 7B | 73.2 | 60.0 | 47.3 | 59.5 | 82.8 | 51.3 | 62.4 |
| 7B–9B | Qwen/Qwen3-8B-Base | 8B | 74.8 | 65.0 | 52.5 | 52.2 | 83.4 | 61.5 | 64.9 |
| 7B–9B | QCRI/Fanar-1-9B-Instruct | 9B | 76.4 | 65.8 | 52.7 | 73.3 | 88.3 | 58.6 | 69.2 |
| 7B–9B | **Hala‑9B** | 9B | 78.3 | 65.6 | 53.8 | 70.4 | 89.6 | 61.4 | **69.9** |
> **Evaluation protocol**: `lighteval` on **ArabicMMLU (OALL‑2)** excluding AlRage.
---
## 📚 Citation
If you find **Hala** useful, please cite:
```bibtex
@misc{hammoud2025halatechnicalreportbuilding,
title={Hala Technical Report: Building Arabic-Centric Instruction & Translation Models at Scale},
author={Hasan Abed Al Kader Hammoud and Mohammad Zbeeb and Bernard Ghanem},
year={2025},
url={https://arxiv.org/abs/2509.14008},
}
``` |