Instructions to use nvidia/Nemotron-3-Labs-Ultra-Math-RL with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nvidia/Nemotron-3-Labs-Ultra-Math-RL with Transformers:

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

pipe = pipeline("text-generation", model="nvidia/Nemotron-3-Labs-Ultra-Math-RL", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nvidia/Nemotron-3-Labs-Ultra-Math-RL", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("nvidia/Nemotron-3-Labs-Ultra-Math-RL", trust_remote_code=True, 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 nvidia/Nemotron-3-Labs-Ultra-Math-RL with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/nvidia/Nemotron-3-Labs-Ultra-Math-RL

SGLang

How to use nvidia/Nemotron-3-Labs-Ultra-Math-RL 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 "nvidia/Nemotron-3-Labs-Ultra-Math-RL" \
    --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": "nvidia/Nemotron-3-Labs-Ultra-Math-RL",
		"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 "nvidia/Nemotron-3-Labs-Ultra-Math-RL" \
        --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": "nvidia/Nemotron-3-Labs-Ultra-Math-RL",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nvidia/Nemotron-3-Labs-Ultra-Math-RL with Docker Model Runner:
```
docker model run hf.co/nvidia/Nemotron-3-Labs-Ultra-Math-RL
```

igitman commited on 4 days ago

Commit

6849d3e

verified ·

1 Parent(s): 48eaf0c

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -28,7 +28,7 @@ track_downloads: true
 Nemotron-3-Labs-Ultra-Math-RL is a decoder-only transformer language model specialized for mathematical reasoning, trained to solve difficult mathematical problems and identify mistakes in proofs, and deployed as part of an ensemble system that achieved a gold-medal level score at the International Mathematical Olympiad 2026.
-Full details can be found at our technical report [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://github.com/NVIDIA-NeMo/Skills/blob/main/recipes/nemotron-imo-tts/paper.pdf).
 Nemotron-3-Labs-Ultra-Math-RL was developed by NVIDIA as a part of Nemotron.
@@ -52,7 +52,7 @@ Researchers and developers focused on AI-driven mathematical reasoning and proof
 ## Reference(s)
-* [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://github.com/NVIDIA-NeMo/Skills/blob/main/recipes/nemotron-imo-tts/paper.pdf) (technical report)
 * [nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16)
 * [nvidia/Nemotron-Math-Proofs-v3-RL](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v3-RL)
 * [nvidia/Nemotron-IMO-Bench](https://huggingface.co/datasets/nvidia/Nemotron-IMO-Bench)
@@ -174,7 +174,7 @@ Context length defaults to 256k above. To use up to 1M, set `VLLM_ALLOW_LONG_MAX
 ### Evaluation Dataset
-**Benchmark Score:** As part of an ensemble system, this model achieved a gold-medal-level score at the International Mathematical Olympiad 2026. Additional evaluation results are available in the [Accompanying Tech Report](https://github.com/NVIDIA-NeMo/Skills/blob/main/recipes/nemotron-imo-tts/paper.pdf).
 **Data Collection Method by dataset:** Hybrid: Automated, manually-collected, Synthetic<br>
 **Labeling Method by dataset:** Hybrid: Automated, manually-labeled, Synthetic<br>

 Nemotron-3-Labs-Ultra-Math-RL is a decoder-only transformer language model specialized for mathematical reasoning, trained to solve difficult mathematical problems and identify mistakes in proofs, and deployed as part of an ensemble system that achieved a gold-medal level score at the International Mathematical Olympiad 2026.
+Full details can be found at our technical report [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://arxiv.org/abs/2609.10712).
 Nemotron-3-Labs-Ultra-Math-RL was developed by NVIDIA as a part of Nemotron.
 ## Reference(s)
+* [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://arxiv.org/abs/2609.10712) (technical report)
 * [nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16)
 * [nvidia/Nemotron-Math-Proofs-v3-RL](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v3-RL)
 * [nvidia/Nemotron-IMO-Bench](https://huggingface.co/datasets/nvidia/Nemotron-IMO-Bench)
 ### Evaluation Dataset
+**Benchmark Score:** As part of an ensemble system, this model achieved a gold-medal-level score at the International Mathematical Olympiad 2026. Additional evaluation results are available in the [Accompanying Tech Report](https://arxiv.org/abs/2609.10712).
 **Data Collection Method by dataset:** Hybrid: Automated, manually-collected, Synthetic<br>
 **Labeling Method by dataset:** Hybrid: Automated, manually-labeled, Synthetic<br>