Instructions to use lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50 with Transformers:

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

pipe = pipeline("text-generation", model="lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50")
model = AutoModelForCausalLM.from_pretrained("lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50")
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 lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50

SGLang

How to use lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50 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 "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50" \
    --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": "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50",
		"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 "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50" \
        --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": "lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50 with Docker Model Runner:
```
docker model run hf.co/lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50
```

qwen3-4B-refiner-rubric-rl-step50

This model is a GRPO-trained checkpoint (step 50) of the Qwen3-4B refiner, fine-tuned with reinforcement learning using per-instance rubric rewards.

Experiment name: instance-rubric-rl-5e-6-answer_only

Training Details

Base model: lihaoxin2020/qwen3-4B-refiner-sft-step-3201
Training method: GRPO (Group Relative Policy Optimization) with DeepSpeed Stage 3
Refiner mode: answer_only
Training script: open_instruct/grpo_fast_refiner_rubric.py

Hyperparameters

Parameter	Value
Learning rate	5e-6
LR scheduler	constant
Beta (KL penalty)	0.001
KL estimator	kl3
Advantage normalization	standard
Samples per prompt (rollout)	8
Unique prompts per rollout	32
Mini batches	1
Epochs per batch	1
Per-device train batch size	1
Temperature	1.0
Seed	42
Async mode	true
Adam offload	true
vLLM sync backend	nccl

Sequence Lengths

Parameter	Value
Max token length	8192
Max prompt token length	6144
Response length	1024
Pack length	8192

Reward Configuration

Parameter	Value
Verification reward	10.0
Non-stop penalty	false
Gate judge score with format bonus	true
Apply paper citation reward	true
Paper citation weight	0.2

Dataset

Training: lihaoxin2020/rl_hard_gpt5_sft (split: train)
Evaluation: lihaoxin2020/rl_hard_gpt5_sft (16 samples, split: train)

Infrastructure

DeepSpeed stage: 3
Learners per node: 1
vLLM engines: 1
vLLM tensor parallel size: 1
vLLM GPU memory utilization: 0.90
Judge model: Qwen/Qwen3.5-35B-A3B (via vLLM)

Downloads last month: 8

Safetensors

Model size

196k params

Tensor type

BF16

Model tree for lihaoxin2020/qwen3-4B-refiner-rubric-rl-step50

Base model

lihaoxin2020/qwen3-4B-refiner-sft-step-3201

Finetuned

(3)

this model