Instructions to use Zigeng/DMax-Math-16B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Zigeng/DMax-Math-16B with Transformers:

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

pipe = pipeline("text-generation", model="Zigeng/DMax-Math-16B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Zigeng/DMax-Math-16B", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Zigeng/DMax-Math-16B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Zigeng/DMax-Math-16B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Zigeng/DMax-Math-16B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Zigeng/DMax-Math-16B

SGLang

How to use Zigeng/DMax-Math-16B 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 "Zigeng/DMax-Math-16B" \
    --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": "Zigeng/DMax-Math-16B",
		"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 "Zigeng/DMax-Math-16B" \
        --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": "Zigeng/DMax-Math-16B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Zigeng/DMax-Math-16B with Docker Model Runner:
```
docker model run hf.co/Zigeng/DMax-Math-16B
```

Zigeng commited on Apr 9

Commit

6aec62a

verified ·

1 Parent(s): ff5c1d3

Update README.md

Browse files

Files changed (1) hide show

README.md +0 -10

README.md CHANGED Viewed

@@ -38,16 +38,6 @@ base_model:
   <em>Superior Parallelism-Accuracy Trade-off, Increased TPF with Maintained Accuracy.</em>
 </div>
-## 💡 Introduction
-We present DMax, a new paradigm for efficient dLLMs. It mitigates error accumulation in parallel decoding, enabling aggressive decoding parallelism while preserving generation quality. Unlike conventional masked dLLMs that decode through a binary mask-to-token transition, DMax reformulates decoding as a progressive self-refinement from mask embeddings to token embeddings. At the core of our approach is On-Policy Uniform Training, a novel training strategy that efficiently unifies masked and uniform dLLMs, equipping the model to recover clean tokens from both masked inputs and its own erroneous predictions. Building on this foundation, we further intoduce Soft Parallel Decoding. Extensive experiments across a variety of benchmarks demonstrate the effectiveness of DMax.
-<!-- ![figure](assets/intro.png) -->
-<div align="center">
-  <img src="assets/train.png" width="100%" />
-  <br>
-  <em>Overview of the On-Policy Uniform Training.</em>
-</div>
 ## 💻 Model and Datasets

   <em>Superior Parallelism-Accuracy Trade-off, Increased TPF with Maintained Accuracy.</em>
 </div>
 ## 💻 Model and Datasets