Instructions to use huihui-ai/DeepSeek-R1-bf16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use huihui-ai/DeepSeek-R1-bf16 with Transformers:

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

pipe = pipeline("text-generation", model="huihui-ai/DeepSeek-R1-bf16", 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("huihui-ai/DeepSeek-R1-bf16", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("huihui-ai/DeepSeek-R1-bf16", trust_remote_code=True)
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use huihui-ai/DeepSeek-R1-bf16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "huihui-ai/DeepSeek-R1-bf16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "huihui-ai/DeepSeek-R1-bf16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/huihui-ai/DeepSeek-R1-bf16

SGLang

How to use huihui-ai/DeepSeek-R1-bf16 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 "huihui-ai/DeepSeek-R1-bf16" \
    --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": "huihui-ai/DeepSeek-R1-bf16",
		"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 "huihui-ai/DeepSeek-R1-bf16" \
        --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": "huihui-ai/DeepSeek-R1-bf16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use huihui-ai/DeepSeek-R1-bf16 with Docker Model Runner:
```
docker model run hf.co/huihui-ai/DeepSeek-R1-bf16
```

Abliteration requests

by Nexesenex - opened Feb 15, 2025

Discussion

Nexesenex

Feb 15, 2025

Hey Huihui,

I have a request for you.
I'm working on a merge project to leverage Llama 3.x 70b smarts, compliance and desalignement.
Could you please abliterate : https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B ?

And if possible, in that order of priority :

https://huggingface.co/migtissera/Tess-R1-Limerick-Llama-3.1-70B (an important model also).

https://huggingface.co/SicariusSicariiStuff/Negative_LLAMA_70B (for basic desalignement)

https://huggingface.co/SentientAGI/Dobby-Unhinged-Llama-3.3-70B (for libertarian desalignement)

Thanks you in any case, I'm using a lot your abliterated models in my merging project with very good results already.

Nexesenex changed discussion title from Abliteration request to Abliteration requests Feb 15, 2025

huihui-ai

Owner Feb 16, 2025

We will try later. Please wait.

Nexesenex

Feb 16, 2025

Thanks guys for considering this, of course I'll wait! :-)

Here's what I achieved so far, only based on merging different abliterated models you made available into a "smart model" for my ulterior merges :

https://huggingface.co/Nexesenex/Llama_3.x_70b_Smarteaz_V1

huihui-ai

Owner Feb 18, 2025

https://huggingface.co/huihui-ai/Tess-R1-Limerick-Llama-3.1-70B-abliterated

Nexesenex

Feb 18, 2025

Thank you folks, thank you !

huihui-ai

Owner Feb 18, 2025

https://huggingface.co/SicariusSicariiStuff/Negative_LLAMA_70B, should not require any ablation. If there is a rejection, you can state the problem that was rejected in this discussion.