Instructions to use deepseek-ai/DeepSeek-V3-0324 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use deepseek-ai/DeepSeek-V3-0324 with Transformers:

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

pipe = pipeline("text-generation", model="deepseek-ai/DeepSeek-V3-0324", 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("deepseek-ai/DeepSeek-V3-0324", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("deepseek-ai/DeepSeek-V3-0324", trust_remote_code=True, 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=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-V3-0324 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "deepseek-ai/DeepSeek-V3-0324"
# 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-V3-0324",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/deepseek-ai/DeepSeek-V3-0324

SGLang

How to use deepseek-ai/DeepSeek-V3-0324 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-V3-0324" \
    --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-V3-0324",
		"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-V3-0324" \
        --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-V3-0324",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use deepseek-ai/DeepSeek-V3-0324 with Docker Model Runner:
```
docker model run hf.co/deepseek-ai/DeepSeek-V3-0324
```

Downloading weights without duplicates

#52

by vadimkantorov - opened Apr 1, 2025

Discussion

vadimkantorov

Apr 1, 2025

•

edited Apr 1, 2025

Downloading with regular git/lfs makes duplicates in .git/lfs/objects which are quite huge for 680Gb weights files:

sudo apt-get install git-lfs
git lfs install

# git clone https://huggingface.co/deepseek-ai/DeepSeek-V3-0324
# du -sh DeepSeek-V3-0324
# # 1.3T    DeepSeek-V3-0324/
# du -sh DeepSeek-V3-0324/.git/lfs
# # 642G    DeepSeek-V3-0324/.git/lfs

How do I download the weights files without any duplication?

Would huggingface_hub.snapshot_download not produce duplicates / any extra cache? (I'm worried of this cache https://huggingface.co/docs/huggingface_hub/en/guides/manage-cache)

pip install hf_transfer huggingface_hub[hf_transfer] 
HF_HUB_ENABLE_HF_TRANSFER=1 python -c 'import huggingface_hub; huggingface_hub.snapshot_download(repo_id="deepseek-ai/DeepSeek-V3-0324",local_dir="deepseek-ai/DeepSeek-V3-0324",allow_patterns=["*.safetensors"])'

CHNtentes

Apr 1, 2025

just use huggingface-cli download deepseek-ai/DeepSeek-V3-0324 should be fine

vadimkantorov

Apr 2, 2025

Thanks! Maybe adding a note about this directly in the README would be very helpful for novices.

As DeepSeek is one of really big open models, so having a warning that git clone https://huggingface.co/deepseek-ai/DeepSeek-V3-0324 would lead to duplicating the 642Gb would be useful, and a command for fast non-duplicating download would be very helpful.

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