Instructions to use AIOR-Research/SOLID-StepORLM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AIOR-Research/SOLID-StepORLM with Transformers:

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

pipe = pipeline("text-generation", model="AIOR-Research/SOLID-StepORLM")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AIOR-Research/SOLID-StepORLM")
model = AutoModelForCausalLM.from_pretrained("AIOR-Research/SOLID-StepORLM", 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 AIOR-Research/SOLID-StepORLM with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/AIOR-Research/SOLID-StepORLM

SGLang

How to use AIOR-Research/SOLID-StepORLM 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 "AIOR-Research/SOLID-StepORLM" \
    --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": "AIOR-Research/SOLID-StepORLM",
		"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 "AIOR-Research/SOLID-StepORLM" \
        --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": "AIOR-Research/SOLID-StepORLM",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AIOR-Research/SOLID-StepORLM with Docker Model Runner:
```
docker model run hf.co/AIOR-Research/SOLID-StepORLM
```

SOLID-StepORLM / README.md

JR-James-0125

Update README.md

cb618db verified 15 days ago

preview code

Raw

History Blame

1.29 kB

	---
	library_name: transformers
	pipeline_tag: text-generation
	base_model: Chenyu-Zhou/StepORLM-Qwen3-8B
	tags:
	- solid-opt
	- operations-research
	- mathematical-optimization
	- self-distillation
	- grpo
	- qwen3
	---

	# SOLID-StepORLM

	This is the checkpoint of SOLID (Solver-Informed Self-Distillation) built from `Chenyu-Zhou/StepORLM-Qwen3-8B` for operations-research modeling and solver-backed answer generation.

	The model was trained with GRPO and solver-informed token-level KL supervision. It uses the COPT-style StepORLM response template.

	## Evaluation

	Each problem was sampled 64 times. `maj@64` is majority-vote accuracy; `pass@k` uses the unbiased pass-at-k estimator.

	\| Dataset \| maj@64 \| pass@1 \| pass@2 \| pass@4 \|
	\|---\|---:\|---:\|---:\|---:\|
	\| OptMATH \| 31.33 \| 18.25 \| 24.40 \| 30.28 \|
	\| MAMO-Complex \| 70.44 \| 66.43 \| 71.58 \| 74.79 \|
	\| InOR \| 48.00 \| 39.81 \| 46.07 \| 50.59 \|



	## Usage

	```python
	from transformers import AutoModelForCausalLM, AutoTokenizer

	model_id = "JR-James-0125/StepORLM-SOLID-opt"
	tokenizer = AutoTokenizer.from_pretrained(model_id)
	model = AutoModelForCausalLM.from_pretrained(
	model_id,
	torch_dtype="auto",
	device_map="auto",
	)
	```

	The generated optimization code expects a compatible COPT environment for execution.