Instructions to use YoussefAhmed26/SmolLM3-NL2Prompt-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use YoussefAhmed26/SmolLM3-NL2Prompt-3B with Transformers:

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

pipe = pipeline("text-generation", model="YoussefAhmed26/SmolLM3-NL2Prompt-3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("YoussefAhmed26/SmolLM3-NL2Prompt-3B")
model = AutoModelForCausalLM.from_pretrained("YoussefAhmed26/SmolLM3-NL2Prompt-3B", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use YoussefAhmed26/SmolLM3-NL2Prompt-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "YoussefAhmed26/SmolLM3-NL2Prompt-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "YoussefAhmed26/SmolLM3-NL2Prompt-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/YoussefAhmed26/SmolLM3-NL2Prompt-3B

SGLang

How to use YoussefAhmed26/SmolLM3-NL2Prompt-3B 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 "YoussefAhmed26/SmolLM3-NL2Prompt-3B" \
    --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": "YoussefAhmed26/SmolLM3-NL2Prompt-3B",
		"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 "YoussefAhmed26/SmolLM3-NL2Prompt-3B" \
        --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": "YoussefAhmed26/SmolLM3-NL2Prompt-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use YoussefAhmed26/SmolLM3-NL2Prompt-3B with Docker Model Runner:
```
docker model run hf.co/YoussefAhmed26/SmolLM3-NL2Prompt-3B
```

Model Card for YoussefAhmed26/SmolLM3-NL2Prompt-3B

A 3B-parameter prompt-generation model built to convert natural-language instructions into structured, high-quality prompts.
This model was developed using curated data generated from multiple frontier LLMs, including ChatGPT-5, Claude Sonnet 4.5, and Gemini 2.5 Pro.

Model Description

SmolLM3-NL2Prompt-3B is a lightweight prompt-generation model designed to take free-form user instructions and transform them into optimized, structured prompts for downstream LLMs.

It was created using a dataset of prompt examples produced by several top-tier AI systems (ChatGPT-5, Claude Sonnet 4.5, and Gemini 2.5 Pro), normalized and aligned using custom formatting rules.

Developed by: Youssef Ahmed
Model type: Prompt-generation LLM
Size: 3B parameters
Language(s): English
License: MIT (or your chosen license)
Finetuned from: SmolLM3 base model
Model ID: YoussefAhmed26/SmolLM3-NL2Prompt-3B

Model Sources

Repository: https://huggingface.co/YoussefAhmed26/SmolLM3-NL2Prompt-3B
Paper: None
Demo: (optional)

Uses

Direct Use

Convert natural language instructions into optimized prompts
Standardize user inputs for agents, chatbots, or pipelines
Improve clarity and structure before querying larger LLMs
Reduce ambiguity in user queries

Downstream Use

Used as a preprocessing module in LLM-based applications
Helps maintain consistent quality in multi-agent systems
Useful for developers building internal tools or automated workflows

Out-of-Scope Use

Factual reasoning
Safety-critical or legal/medical uses
Attempts to imitate proprietary models used during prompt collection

Bias, Risks, and Limitations

Inherits stylistic preferences from ChatGPT, Claude, and Gemini
May generate overly structured or overly formal prompts
Not responsible for fact-checking or content correctness
Outputs depend on user input clarity

Recommendations

Always review generated prompts before using them
Use additional filtering in production systems
Ensure compliance with the terms of the LLMs whose outputs were used for building the dataset

How to Get Started

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "YoussefAhmed26/SmolLM3-NL2Prompt-3B"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

input_text = "Write a prompt for summarizing a scientific article."
inputs = tokenizer(input_text, return_tensors="pt")

outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Downloads last month: 4

Safetensors

Model size

3B params

Tensor type

F16