Instructions to use alamios/Mistral-Small-3.1-DRAFT-0.5B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use alamios/Mistral-Small-3.1-DRAFT-0.5B with Transformers:

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

pipe = pipeline("text-generation", model="alamios/Mistral-Small-3.1-DRAFT-0.5B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("alamios/Mistral-Small-3.1-DRAFT-0.5B")
model = AutoModelForCausalLM.from_pretrained("alamios/Mistral-Small-3.1-DRAFT-0.5B", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use alamios/Mistral-Small-3.1-DRAFT-0.5B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "alamios/Mistral-Small-3.1-DRAFT-0.5B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "alamios/Mistral-Small-3.1-DRAFT-0.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/alamios/Mistral-Small-3.1-DRAFT-0.5B

SGLang

How to use alamios/Mistral-Small-3.1-DRAFT-0.5B 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 "alamios/Mistral-Small-3.1-DRAFT-0.5B" \
    --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": "alamios/Mistral-Small-3.1-DRAFT-0.5B",
		"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 "alamios/Mistral-Small-3.1-DRAFT-0.5B" \
        --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": "alamios/Mistral-Small-3.1-DRAFT-0.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use alamios/Mistral-Small-3.1-DRAFT-0.5B with Docker Model Runner:
```
docker model run hf.co/alamios/Mistral-Small-3.1-DRAFT-0.5B
```

Works with SGLang Standalone. Details on how to build these models?

by anttip - opened Oct 29, 2025

Discussion

anttip

Oct 29, 2025

SGLang implemented a new speculative decoding method "Standalone", built on Eagle, which supports this model: https://github.com/sgl-project/sglang/pull/10090

Standalone can run this speculator with MistralSmall models, reducing latencies quite a lot. But the Qwen model this speculator was built on doesn't tensor-parallelize to 4GPU or 8-GPU instances, which limits the options for deploying this speculator.

Are there more details on how this speculator was built, so it can be done on other base-models?

alamios

Owner Nov 2, 2025

It was built by:

Transplanting vocab of Mistral into Qwen, making Qwenstral (https://huggingface.co/alamios/Qwenstral-Small-3.1-0.5B)
Then generating a small dataset of Mistral responses on multiple tasks and datasets (https://huggingface.co/datasets/alamios/Mistral-Small-24B-Instruct-2501-Conversations)
Finally, SFT'ing Qwenstral on Mistral's outputs to align them (superficially, on common tasks)

I'm not familiar with that problem. What exactly limits Qwen for parallelization and what small models are suitable for that?

anttip

Nov 2, 2025

Thanks for the details. I found the repo and other models, and built some similar speculators.

To tensor-parallelize a speculator, the dimensions and sizes have to be divisible by the tensor parallel number. So with 4 or 8, the values of the model have to divide by that. This model won't do with tp>2. The model has num_attention_heads 14, which divides by 2, but not by 4.

The one ok-quality model that does is llama3.2-1b, which can be transplanted and does tensor-parallelize to 4. But SGLang still throws warnings with tp>2, and there's no speedups with tp>2 for some reason.

alamios

Owner Nov 4, 2025

Interesting, I haven't realized that, I'll take it into account if I return to creating draft models.

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