Instructions to use GSAI-ML/LLaDA-8B-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GSAI-ML/LLaDA-8B-Base with Transformers:

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

pipe = pipeline("text-generation", model="GSAI-ML/LLaDA-8B-Base", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("GSAI-ML/LLaDA-8B-Base", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use GSAI-ML/LLaDA-8B-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GSAI-ML/LLaDA-8B-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GSAI-ML/LLaDA-8B-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/GSAI-ML/LLaDA-8B-Base

SGLang

How to use GSAI-ML/LLaDA-8B-Base 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 "GSAI-ML/LLaDA-8B-Base" \
    --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": "GSAI-ML/LLaDA-8B-Base",
		"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 "GSAI-ML/LLaDA-8B-Base" \
        --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": "GSAI-ML/LLaDA-8B-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use GSAI-ML/LLaDA-8B-Base with Docker Model Runner:
```
docker model run hf.co/GSAI-ML/LLaDA-8B-Base
```

weight tying

by spawn99 - opened Feb 27, 2025

Discussion

spawn99

Feb 27, 2025

The paper describes LLaDA as a 'vanilla Transformer' predicting masked tokens, and my inference code expects logits over the vocab at each diffusion step. Given model.transformer.ff_out.weight exists but lm_head.weight doesn’t, is this ff_out the output projection for LLaDA?
Is there weight tying going on, unlike the 'vanilla' description in the paper?

nieshen

GSAI-ML org Mar 4, 2025

Thank you for your attention!

The LLaDA model is derived from removing the causal mask from an autoregressive model (as detailed in https://github.com/ML-GSAI/LLaDA/blob/main/GUIDELINES.md), so we call it a vanilla Transformer. By using this term, we emphasize that, unlike other diffusion language models, LLaDA does not require inputting the timestep into the network. I'm not sure if I've fully understood your question. More discussions are welcome.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment