Instructions to use yifanyu/QwenDLM-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use yifanyu/QwenDLM-4B with Transformers:

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

pipe = pipeline("text-generation", model="yifanyu/QwenDLM-4B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("yifanyu/QwenDLM-4B", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use yifanyu/QwenDLM-4B with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/yifanyu/QwenDLM-4B

SGLang

How to use yifanyu/QwenDLM-4B 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 "yifanyu/QwenDLM-4B" \
    --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": "yifanyu/QwenDLM-4B",
		"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 "yifanyu/QwenDLM-4B" \
        --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": "yifanyu/QwenDLM-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use yifanyu/QwenDLM-4B with Docker Model Runner:
```
docker model run hf.co/yifanyu/QwenDLM-4B
```

QwenDLM-4B

QwenDLM (4B): Qwen3-4B-Thinking-2507 converted into a block diffusion language model (SDAR-style, 256-token blocks) with a self-conditioning residual adapter, the start model of QwenDLM-4B-ORL

Results

Accuracy (%) / tokens per forward (TPF); each model uses its selected decoder.

Model	GSM8K	MATH-500	HumanEval	MBPP
QwenDLM-4B	90.4 / 2.4	88.0 / 2.1	82.9 / 1.8	79.2 / 2.1
QwenDLM-4B-ORL	92.4 / 5.9	92.2 / 5.8	89.6 / 11.1	79.4 / 8.0

Speed

Tokens per second for a single request on one H100 (BF16), with the paper's prompts and decoders. dlmrt decodes QwenDLM-4B about 1.5× faster than SGLang, and ORL raises it to 1,298 tokens/s:

Model	Engine	GSM8K	HumanEval	Mean
QwenDLM-4B	SGLang	210	164	187
QwenDLM-4B	dlmrt	328	249	289
QwenDLM-4B-ORL	SGLang	718	1,148	933
QwenDLM-4B-ORL	dlmrt	987	1,609	1,298

Usage

Plain transformers generation does not run the paper's decoding. Two engines do: dlmrt, our inference engine, and the ORL repository's SGLang engine and decoder configuration, which the paper's evaluation uses.

dlmrt

dlmrt, our inference engine, decodes one request at a time with the paper's decoder (role="base") and measures the paper's throughput; see its README for installation, tensor parallelism and the commands of every paper cell.

from dlmrt import LLM

llm = LLM("yifanyu/QwenDLM-4B", role="base")
out = llm.generate("Write a Python function that reverses a linked list.", max_new_tokens=4096)
print(out.text)
print(f"{out.request_tps:.0f} tokens/s")

python -m dlmrt serve --model yifanyu/QwenDLM-4B --role base --port 18080
curl -N localhost:18080/generate -d '{"messages": [{"role": "user", "content": "Hi!"}], "max_new_tokens": 512}'

The paper's evaluation pipeline

Reproduce the paper's numbers (GSM8K, MATH-500, HumanEval, MBPP) with the repository's evaluation pipeline:

git clone https://github.com/Introspective-Diffusion/ORL orl && cd orl/evaluation
export ORL_EVAL_WORK=/path/to/eval-work
bash scripts/build_images.sh && bash scripts/prepare.sh
hf download yifanyu/QwenDLM-4B --local-dir $ORL_EVAL_WORK/hf/QwenDLM-4B
bash scripts/main_table.sh qwen-dlm-base $ORL_EVAL_WORK/hf/QwenDLM-4B

Serve it with the engine settings of the evaluation (paper_decoding.json) and a paper decoder (here GSM8K's):

python3 -m repro.paper.render_config --suite main --model qwen-dlm-base --benchmark gsm8k --seed 1234 --out decoder.yaml
apptainer exec --nv -B "$PWD,$ORL_EVAL_WORK" $ORL_EVAL_WORK/images/sglang-gen.sif \
  env PYTHONPATH=$ORL_EVAL_WORK/runtime/qwen-dlm/engine/inference/sglang \
      SGLANG_FLASHINFER_PREFILL_BACKEND=fa3 SDAR_FUSED_ROW_STATS=1 SDAR_COMMIT_GRAPH=1 \
  python3 -m sglang.launch_server --model-path $ORL_EVAL_WORK/hf/QwenDLM-4B --trust-remote-code \
    --tp-size 1 --max-running-requests 1 --context-length 65536 --attention-backend flashinfer \
    --chunked-prefill-size -1 --sdar-rollout-cuda-graph --sdar-rollout-cuda-graph-capture-sizes 1 \
    --dllm-algorithm SDARPrefixKV --dllm-algorithm-config decoder.yaml --port 30000

Training ORL from this model

This is checkpoint 38750 of the QwenDLM conversion run. The ORL runs start from it after rescaling the residual adapter (the result's weights and configuration are byte-identical to the paper's start model); the training data is the dataset yifanyu/ORL-training-data:

hf download yifanyu/ORL-training-data --repo-type dataset --local-dir $DATA
apptainer exec -B $PWD $SIF $PY training/tools/make_scrp_base.py QwenDLM-4B QwenDLM-4B-scrp8 --gamma 8
bash training/scripts/reproduce_qwen_dlm.sh $RUNS/qwen_dlm QwenDLM-4B-scrp8 $DATA --sif $SIF --python $PY

Checkpoint identity

Paper role qwen-dlm-base. checkpoint_manifest.json records the hashes of the weights, tokenizer, configuration and custom model code, and paper_decoding.json the engine and benchmark-specific decoding settings, which downloading the model does not apply.

Related Models

Model	Hugging Face	Description
Qwen3-4B-Thinking-2507	Qwen/Qwen3-4B-Thinking-2507	Start model of this checkpoint
QwenDLM-4B	yifanyu/QwenDLM-4B	This model
QwenDLM-4B-ORL	yifanyu/QwenDLM-4B-ORL	ORL on QwenDLM-4B
LLaDA2.1-mini-ORL	yifanyu/LLaDA2.1-mini-ORL	ORL on LLaDA-2.1-mini
LLaDA2.1-mini-ORL-NVFP4	yifanyu/LLaDA2.1-mini-ORL-NVFP4	NVFP4 quantization of LLaDA2.1-mini-ORL
DiffusionGemma-26B-A4B-it-ORL	yifanyu/DiffusionGemma-26B-A4B-it-ORL	ORL on DiffusionGemma-26B-A4B
DiffusionGemma-26B-A4B-it-ORL-NVFP4	yifanyu/DiffusionGemma-26B-A4B-it-ORL-NVFP4	NVFP4 quantization of DiffusionGemma-26B-A4B-it-ORL
ORL-training-data	yifanyu/ORL-training-data	Dataset of the three ORL training runs

Citation

@misc{yu2026orl,
  title         = {On-Policy Residual Learning for Diffusion Language Models},
  author        = {Yu, Yifan and Jian, Yuqing and Li, Jisen and Hu, Yuezhou and Zhang, Tianyi and Wu, Xiaoxia and
                   Zhu, Rui-Jie and Wang, Junxiong and Zhou, Zhongzhu and Lai, Fan and Zou, James and Xu, Chenfeng},
  year          = {2026},
  eprint        = {2610.20687},
  archivePrefix = {arXiv},
  primaryClass  = {cs.LG},
  url           = {https://arxiv.org/abs/2610.20687}
}