Instructions to use unsloth/DeepSeek-R1-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use unsloth/DeepSeek-R1-GGUF with Transformers:

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

pipe = pipeline("text-generation", model="unsloth/DeepSeek-R1-GGUF", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("unsloth/DeepSeek-R1-GGUF", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("unsloth/DeepSeek-R1-GGUF", trust_remote_code=True)

llama-cpp-python

How to use unsloth/DeepSeek-R1-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="unsloth/DeepSeek-R1-GGUF",
	filename="DeepSeek-R1-BF16/DeepSeek-R1.BF16-00001-of-00030.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use unsloth/DeepSeek-R1-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf unsloth/DeepSeek-R1-GGUF:Q4_K_M

Use Docker

docker model run hf.co/unsloth/DeepSeek-R1-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use unsloth/DeepSeek-R1-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "unsloth/DeepSeek-R1-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "unsloth/DeepSeek-R1-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/unsloth/DeepSeek-R1-GGUF:Q4_K_M

SGLang

How to use unsloth/DeepSeek-R1-GGUF 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 "unsloth/DeepSeek-R1-GGUF" \
    --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": "unsloth/DeepSeek-R1-GGUF",
		"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 "unsloth/DeepSeek-R1-GGUF" \
        --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": "unsloth/DeepSeek-R1-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use unsloth/DeepSeek-R1-GGUF with Ollama:
```
ollama run hf.co/unsloth/DeepSeek-R1-GGUF:Q4_K_M
```

Unsloth Studio new

How to use unsloth/DeepSeek-R1-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for unsloth/DeepSeek-R1-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for unsloth/DeepSeek-R1-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for unsloth/DeepSeek-R1-GGUF to start chatting

Docker Model Runner
How to use unsloth/DeepSeek-R1-GGUF with Docker Model Runner:
```
docker model run hf.co/unsloth/DeepSeek-R1-GGUF:Q4_K_M
```

Lemonade

How to use unsloth/DeepSeek-R1-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull unsloth/DeepSeek-R1-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.DeepSeek-R1-GGUF-Q4_K_M

List all available models

lemonade list

Q2_K_XL model is the best? IQ2_XXS is better than Q2_K_XL in mmlu-pro benchmark

#36

by albertchow - opened Feb 23, 2025

Discussion

albertchow

Feb 23, 2025

YudingOOM

Feb 23, 2025

where can we get the mmlu-pro benchmark for Q2_K_XL and IQ2_XXS?

Rotating

Feb 23, 2025

•

edited Feb 23, 2025

Without the data we can't tell. It could just be randomness because we're running R1 at 0.6 temp and it's doing reasoning.

albertchow

Feb 23, 2025

•

edited Feb 27, 2025

well, I only tested computer science (410 questions) with zero-shot by following usage recommendations mentioned in DeepSeek-R1 page, the score of Q2_K_XL is 80.24, and the score of IQ2_XXS is 84.15.
the test time is very long, therefore, I only tested once for each quant.

the following is the example of the first question and the corresponding response:

Rotating

Feb 24, 2025

well, I only tested computer science (410 questions) with zero-shot by following usage recommendations mentioned in DeepSeek-R1 page, the score of Q2_K_XL is 80.24, and the score of IQ2_XXS is 84.15.
the test time is very long, therefore, I only tested once for each quant.

I remember something similar happening consistently back in the day on some old model where a lower quant somehow got one question right where the q8 didn't.

albertchow

Feb 24, 2025

•

edited Feb 24, 2025

I remember something similar happening consistently back in the day on some old model where a lower quant somehow got one question right where the q8 didn't.

The MMLU-Pro computer science subject has 410 questions, so a few outliers have little impact on the final scores.

Rotating

Feb 24, 2025

I remember something similar happening consistently back in the day on some old model where a lower quant somehow got one question right where the q8 didn't.

The MMLU-Pro computer science subject has 410 questions, so a few outliers have little impact on the final scores.

It's interesting for sure. For what it's worth, the perplexity scores don't seem to indicate a problem with the quant though. https://huggingface.co/unsloth/DeepSeek-R1-GGUF/discussions/37

albertchow

Feb 24, 2025

•

edited Feb 24, 2025

well, I test R1 distill Qwen-32B model on MMLU-Pro computer science by repeating 8 times, and you can see Q3_K_M even has large model size and lower perplexity, the score is still worse than IQ3_XS. (the GGUF models provided by bartowski)

Rotating

Feb 24, 2025

well, I test R1 distill Qwen-32B model on MMLU-Pro computer science by repeating 8 times, and you can see Q3_K_M even has large model size and lower perplexity, the score is still worse than IQ3_XS. (the GGUF models provided by bartowski)

The IQ quant being better isn't a surprise, and the perplexity difference is within the margin of uncertainty. Having said all that, if I hadn't already downloaded q2_k_xl I might've gone for the 2xxs to be faster and at least in one test be better at computer science. I wonder what IQ3_M would get on mmlu pro computer science.

Rotating

Feb 26, 2025

It turns out xxs is using imatrix and q2_k_xl isn't. That might be the factor involved.

albertchow

Feb 26, 2025

•

edited Feb 26, 2025

As for IQ3_M provided by bartowski, I cannot use max_tokens = 8192 as the smaller models (e.g., IQ2_XXS) did because out of memory, thus, the comparison of IQ3_M is a bit of unfair.
The score of IQ3_M is around 80 (still running, unfinished) under max_tokens = 6300
Therefore, I would like to say the IQ2_XXS provided by unsloth is pretty good.

shimmyshimmer

Unsloth AI org Feb 27, 2025

As for IQ3_M provided by bartowski, I cannot use max_tokens = 8192 as the smaller models (e.g., IQ2_XXS) did because out of memory, thus, the comparison of IQ3_M is a bit of unfair.
The score of IQ3_M is around 80 (still running, unfinished) under max_tokens = 6300
Therefore, I would like to say the IQ2_XXS provided by unsloth is pretty good.

Great stuff thanks for confirming! :)

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