Instructions to use mariklolik/AraToken-Qwen3-0.6B-CPT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mariklolik/AraToken-Qwen3-0.6B-CPT with Transformers:

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

pipe = pipeline("text-generation", model="mariklolik/AraToken-Qwen3-0.6B-CPT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mariklolik/AraToken-Qwen3-0.6B-CPT")
model = AutoModelForCausalLM.from_pretrained("mariklolik/AraToken-Qwen3-0.6B-CPT", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mariklolik/AraToken-Qwen3-0.6B-CPT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mariklolik/AraToken-Qwen3-0.6B-CPT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mariklolik/AraToken-Qwen3-0.6B-CPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mariklolik/AraToken-Qwen3-0.6B-CPT

SGLang

How to use mariklolik/AraToken-Qwen3-0.6B-CPT 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 "mariklolik/AraToken-Qwen3-0.6B-CPT" \
    --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": "mariklolik/AraToken-Qwen3-0.6B-CPT",
		"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 "mariklolik/AraToken-Qwen3-0.6B-CPT" \
        --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": "mariklolik/AraToken-Qwen3-0.6B-CPT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mariklolik/AraToken-Qwen3-0.6B-CPT with Docker Model Runner:
```
docker model run hf.co/mariklolik/AraToken-Qwen3-0.6B-CPT
```

AraToken-Qwen3-0.6B-CPT

The continued-pretraining (CPT) baseline of the AraToken paper. Qwen3-0.6B-Base keeps its original tokenizer and is trained on the same 500M-token FineWeb2-HQ Arabic budget as mariklolik/AraToken-Qwen3-0.6B-LEP: 5,086 steps, transformer layers 24–27 trainable.

Paper: AraToken: Optimizing Arabic Tokenization with Normalization Pipeline and Language Extension for Qwen3. Code: https://github.com/mariklolik/Aratoken.

Model	Arabic BPC ↓
Qwen3-0.6B-Base	1.5446
this model	1.4047
AraToken-Qwen3-0.6B-LEP	1.3219

BPC is bits per character on the first 1,500 documents of the held-out test split of mariklolik/AraToken-FineWeb2-HQ-ar. It compares models with different vocabularies on the same characters.

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "mariklolik/AraToken-Qwen3-0.6B-CPT"
tokenizer = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype="bfloat16")

The tokenizer carries the AraToken normalizer: NFKC, Tatweel removal, Western digits, Latin punctuation and diacritics removal. Raw Arabic text can therefore be passed to it directly.

Citation

@article{kashirskiy2025aratoken,
  title   = {AraToken: Optimizing Arabic Tokenization with Normalization Pipeline and Language Extension for Qwen3},
  author  = {Kashirskiy, Mark and Lipinski, Artiom and Makarov, Ilya},
  journal = {arXiv preprint arXiv:2512.18399},
  year    = {2025}
}

Downloads last month: 330

Safetensors

Model size

0.6B params

Tensor type

BF16

Model tree for mariklolik/AraToken-Qwen3-0.6B-CPT

Base model

Qwen/Qwen3-0.6B-Base

Finetuned

(724)

this model

Dataset used to train mariklolik/AraToken-Qwen3-0.6B-CPT

Paper for mariklolik/AraToken-Qwen3-0.6B-CPT

AraToken: Optimizing Arabic Tokenization with Normalization Pipeline and Language Extension for Qwen3

Paper • 2512.18399 • Published Dec 20, 2025