Instructions to use Simplified-Reasoning/SU-01 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Simplified-Reasoning/SU-01 with Transformers:

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

pipe = pipeline("text-generation", model="Simplified-Reasoning/SU-01")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Simplified-Reasoning/SU-01")
model = AutoModelForCausalLM.from_pretrained("Simplified-Reasoning/SU-01", 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 Simplified-Reasoning/SU-01 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Simplified-Reasoning/SU-01"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Simplified-Reasoning/SU-01",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Simplified-Reasoning/SU-01

SGLang

How to use Simplified-Reasoning/SU-01 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 "Simplified-Reasoning/SU-01" \
    --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": "Simplified-Reasoning/SU-01",
		"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 "Simplified-Reasoning/SU-01" \
        --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": "Simplified-Reasoning/SU-01",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Simplified-Reasoning/SU-01 with Docker Model Runner:
```
docker model run hf.co/Simplified-Reasoning/SU-01
```

rzzhan commited on May 20

Commit

edbca7a

verified ·

1 Parent(s): a3509c4

chore: add shared decode prompt template

Browse files

Files changed (1) hide show

README.md +11 -0

README.md CHANGED Viewed

@@ -181,6 +181,17 @@ The released TTS implementation is in [`su01-eval/decode`](https://github.com/Si
   <img src="https://github.com/Simplified-Reasoning/SU-01/raw/main/page/source_png/tts_action_length_distribution_1.png" alt="Test-time scaling action length distribution" width="80%">
 </p>
 ---
 <a id="evaluation"></a>

   <img src="https://github.com/Simplified-Reasoning/SU-01/raw/main/page/source_png/tts_action_length_distribution_1.png" alt="Test-time scaling action length distribution" width="80%">
 </p>
+### SU-01 Prompt Template (Direct Decoding)
+Rendered for readability, the default prompt is:
+```text
+Please solve the following mathematical olympiad problem. Show your complete reasoning and proof.
+1. Please use LaTeX format to represent the variables and formulas used in the solution process and results.
+2. If the problem asks you to find specific values, please put the final answer(s) in \boxed{}.
+3. If the problem requires a proof, present a clear and rigorous argument.
+```
 ---
 <a id="evaluation"></a>