Instructions to use deepseek-ai/DeepSeek-R1-Distill-Qwen-32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use deepseek-ai/DeepSeek-R1-Distill-Qwen-32B with Transformers:

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

pipe = pipeline("text-generation", model="deepseek-ai/DeepSeek-R1-Distill-Qwen-32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("deepseek-ai/DeepSeek-R1-Distill-Qwen-32B")
model = AutoModelForCausalLM.from_pretrained("deepseek-ai/DeepSeek-R1-Distill-Qwen-32B")
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]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use deepseek-ai/DeepSeek-R1-Distill-Qwen-32B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B

SGLang

How to use deepseek-ai/DeepSeek-R1-Distill-Qwen-32B 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 "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B" \
    --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": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",
		"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 "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B" \
        --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": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use deepseek-ai/DeepSeek-R1-Distill-Qwen-32B with Docker Model Runner:
```
docker model run hf.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
```

The input starts with the token "<|begin▁of▁sentence|>" repeated twice. / 输入开头重复2次“<|begin▁of▁sentence|>”

#31

by Gya123 - opened Feb 5, 2025

Discussion

Gya123

Feb 5, 2025

I have two questions:

In the configuration file "tokenizer_config.json" for the Qwen series model, the "tokenizer_class" is set to "LlamaTokenizerFast." I'm not sure why this is the case, but after testing, the results are consistent with those of QwenTokenizer.
In the "tokenizer_config.json," "add_bos_token" is set to true, meaning that the tokenizer will automatically add a bos_token, which is "<|begin▁of▁sentence|>". However, when using tokenizer.apply_chat_template, it also adds "<|begin▁of▁sentence|>", resulting in the final output starting with two repeated "<|begin▁of▁sentence|>" tokens.

Here is the reproduction code:

prompt = "计算1+1"
messages = [
    {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

input_ids = model_inputs['input_ids']
tokenizer.decode(input_ids[0])

# output:
# <｜begin▁of▁sentence｜><｜begin▁of▁sentence｜>You are Qwen, created by Alibaba Cloud. You are a helpful assistant.<｜User｜>计算1+1<｜Assistant｜>

我有2个问题：
1、在qwen系列模型的配置文件 "tokenizer_config.json" 中，"tokenizer_class"设置为"LlamaTokenizerFast"，这个不知道是为什么，不过测下来和QwenTokenizer的返回结果是一致的；

2、tokenizer_config.json中 "add_bos_token": true, 也就是tokenizer时会自动添加bos_token，也就是“<｜begin▁of▁sentence｜>”，但是tokenizer.apply_chat_template时也会添加“<｜begin▁of▁sentence｜>”，也就导致最终开头是2个重复的“<｜begin▁of▁sentence｜>”

请问以上是否符合预期（训练时也一样输入2次<｜begin▁of▁sentence｜>），如果一致则应该不需要改动。

dongxx1104

Jul 11, 2025

same issue

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