Instructions to use diffusion-reasoning/gdsd_countdown_dream with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use diffusion-reasoning/gdsd_countdown_dream with Transformers:

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

pipe = pipeline("text-generation", model="diffusion-reasoning/gdsd_countdown_dream", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("diffusion-reasoning/gdsd_countdown_dream", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use diffusion-reasoning/gdsd_countdown_dream with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/diffusion-reasoning/gdsd_countdown_dream

SGLang

How to use diffusion-reasoning/gdsd_countdown_dream 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 "diffusion-reasoning/gdsd_countdown_dream" \
    --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": "diffusion-reasoning/gdsd_countdown_dream",
		"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 "diffusion-reasoning/gdsd_countdown_dream" \
        --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": "diffusion-reasoning/gdsd_countdown_dream",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use diffusion-reasoning/gdsd_countdown_dream with Docker Model Runner:
```
docker model run hf.co/diffusion-reasoning/gdsd_countdown_dream
```

gdsd_countdown_dream / README.md

jky594176

Add model card for GDSD (#1)

9acd1d0 about 17 hours ago

preview code

raw

history blame contribute delete

1.55 kB

metadata

pipeline_tag: text-generation
library_name: transformers

GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models

This repository contains the model weights for GDSD, as presented in the paper GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models.

Guided Denoiser Self-Distillation (GDSD) is a reinforcement learning framework for diffusion language models (dLLMs). It improves the denoiser of dLLMs by distilling from an advantage-guided self-teacher, bypassing the biases associated with evidence lower bound (ELBO) surrogates used in prior methods. GDSD provides a more stable and effective RL procedure, achieving significant performance gains on planning, math, and coding benchmarks.

Resources

Paper: GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models
GitHub Repository: https://github.com/GaryBall/GDSD

Citation

@misc{tang2026gdsdreinforcementlearningguided,
      title={GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models}, 
      author={Xiaohang Tang and Keyue Jiang and Che Liu and Qifang Zhao and Xiaoxiao Xu and Sangwoong Yoon and Ilija Bogunovic},
      year={2026},
      eprint={2605.29398},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2605.29398}, 
}