Instructions to use nvidia/OpenReasoning-Nemotron-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nvidia/OpenReasoning-Nemotron-32B with Transformers:

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

pipe = pipeline("text-generation", model="nvidia/OpenReasoning-Nemotron-32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nvidia/OpenReasoning-Nemotron-32B")
model = AutoModelForCausalLM.from_pretrained("nvidia/OpenReasoning-Nemotron-32B")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use nvidia/OpenReasoning-Nemotron-32B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nvidia/OpenReasoning-Nemotron-32B"
# 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/OpenReasoning-Nemotron-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/nvidia/OpenReasoning-Nemotron-32B

SGLang

How to use nvidia/OpenReasoning-Nemotron-32B 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/OpenReasoning-Nemotron-32B" \
    --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/OpenReasoning-Nemotron-32B",
		"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/OpenReasoning-Nemotron-32B" \
        --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/OpenReasoning-Nemotron-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nvidia/OpenReasoning-Nemotron-32B with Docker Model Runner:
```
docker model run hf.co/nvidia/OpenReasoning-Nemotron-32B
```

Update generation_config.json

by qgallouedec HF Staff - opened Jul 23, 2025

base: refs/heads/main

←

from: refs/pr/4

Discussion Files changed

-3

Files changed (3) hide show

config.json +1 -1
generation_config.json +1 -1
tokenizer_config.json +1 -1

config.json CHANGED Viewed

@@ -5,7 +5,7 @@
   ],
   "attention_dropout": 0.0,
   "bos_token_id": 151643,
-  "eos_token_id": 151643,
   "hidden_act": "silu",
   "hidden_size": 5120,
   "initializer_range": 0.02,

   ],
   "attention_dropout": 0.0,
   "bos_token_id": 151643,
+  "eos_token_id": 151645,
   "hidden_act": "silu",
   "hidden_size": 5120,
   "initializer_range": 0.02,

generation_config.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "_from_model_config": true,
   "bos_token_id": 151643,
-  "eos_token_id": 151643,
   "transformers_version": "4.47.1"
 }

 {
   "_from_model_config": true,
   "bos_token_id": 151643,
+  "eos_token_id": 151645,
   "transformers_version": "4.47.1"
 }

tokenizer_config.json CHANGED Viewed

@@ -197,7 +197,7 @@
   "bos_token": null,
   "chat_template": "{%- if messages[0]['role'] == 'system' %}\n    {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}\n{%- else %}\n    {{- '<|im_start|>system\n<|im_end|>\n' }}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == 'user') or (message.role == 'system' and not loop.first) or (message.role == 'assistant') %}\n        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\n' }}\n{%- endif %}",
   "clean_up_tokenization_spaces": false,
-  "eos_token": "<|endoftext|>",
   "errors": "replace",
   "extra_special_tokens": {},
   "model_max_length": 131072,

   "bos_token": null,
   "chat_template": "{%- if messages[0]['role'] == 'system' %}\n    {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}\n{%- else %}\n    {{- '<|im_start|>system\n<|im_end|>\n' }}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == 'user') or (message.role == 'system' and not loop.first) or (message.role == 'assistant') %}\n        {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\n' }}\n{%- endif %}",
   "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
   "errors": "replace",
   "extra_special_tokens": {},
   "model_max_length": 131072,