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
```

JR-James-0125 commited on 12 days ago

Commit

4f58aa8

verified ·

1 Parent(s): 03cb8a5

Label StepORLM checkpoint as SOLID-opt

Browse files

Files changed (1) hide show

README.md +5 -3

README.md CHANGED Viewed

@@ -3,6 +3,7 @@ library_name: transformers
 pipeline_tag: text-generation
 base_model: Chenyu-Zhou/StepORLM-Qwen3-8B
 tags:
   - operations-research
   - mathematical-optimization
   - self-distillation
@@ -10,7 +11,7 @@ tags:
   - qwen3
 ---
-# SOLID-StepORLM-Qwen3-8B
 This is the step-125 checkpoint of **SOLID (Solver-Informed Self-Distillation)** built from `Chenyu-Zhou/StepORLM-Qwen3-8B` for operations-research modeling and solver-backed answer generation.
@@ -26,12 +27,14 @@ Each problem was sampled 64 times. `maj@64` is majority-vote accuracy; `pass@k`
 | 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 = "JamesX421/SOLID-StepORLM-Qwen3-8B"
 tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
@@ -41,4 +44,3 @@ model = AutoModelForCausalLM.from_pretrained(
 ```
 The generated optimization code expects a compatible COPT environment for execution.

 pipeline_tag: text-generation
 base_model: Chenyu-Zhou/StepORLM-Qwen3-8B
 tags:
+  - solid-opt
   - operations-research
   - mathematical-optimization
   - self-distillation
   - qwen3
 ---
+# StepORLM-SOLID-opt
 This is the step-125 checkpoint of **SOLID (Solver-Informed Self-Distillation)** built from `Chenyu-Zhou/StepORLM-Qwen3-8B` for operations-research modeling and solver-backed answer generation.
 | MAMO-Complex | 70.44 | 66.43 | 71.58 | 74.79 |
 | InOR | 48.00 | 39.81 | 46.07 | 50.59 |
+Across the coherent 697-problem evaluation suite, aggregate pass@1 is **60.16%**.
 ## 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,
 ```
 The generated optimization code expects a compatible COPT environment for execution.