deepseek-ai
/

DeepSeek-V3-0324

Text Generation

text-generation-inference

Model card Files Files and versions

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)
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

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

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
```

Resources

View closed (17)

TemporalMesh Transformer: 29.4 PPL at 48% compute — beats Mamba, new open-source architecture

#95 opened 3 days ago by

Add EvalEval community eval results

#94 opened 5 days ago by

Fix chat_template crash when assistant message omits the `content` key

#93 opened about 1 month ago by

Install & run deepseek-ai/DeepSeek-V3-0324 easily using llmpm

#92 opened 3 months ago by

Add MMLU-Pro evaluation result (81.2)

#91 opened 4 months ago by

What are the original sampling parameters?

#90 opened 6 months ago by

Upload ATS-NSE-Stock-Suite-main.zip

#89 opened 6 months ago by

mobil uygulama

#86 opened 11 months ago by

Update README.md

#85 opened 12 months ago by

Update README.md

#83 opened 12 months ago by

Request: DOI

#82 opened about 1 year ago by

Update README.md

#81 opened about 1 year ago by

Improve model card with paper link and pipeline tag

#80 opened about 1 year ago by

Social Face

#79 opened about 1 year ago by

Fix broken chat formatting with chat_template. Possible fix to degraded experiences across inference providers

#78 opened about 1 year ago by

harrisonvanderbyl

Upload TypeFlow Logo.png

#77 opened about 1 year ago by

Create old ranch

#76 opened about 1 year ago by

Uauai

#75 opened about 1 year ago by

Memory of GPU

#74 opened about 1 year ago by

Update README.md

#73 opened about 1 year ago by

Rename README.md to creat ppt about mosalah player

#72 opened about 1 year ago by

Review of DeepSeek V3 0324

#71 opened about 1 year ago by

henrik.hellstrom76@protonmail.com

#70 opened about 1 year ago by

World's Largest Dataset

#69 opened about 1 year ago by deleted

Feature Request: Can we have a QAT version?

#68 opened about 1 year ago by

Upload Dockerfile

#67 opened about 1 year ago by

Upload ImportedPhoto.765838358.341873.jpeg

#66 opened about 1 year ago by

Upload IMG_1859.png

#64 opened about 1 year ago by

Rename README.md to README.mddocker run -it -p 7860:7860 --platform=linux/amd64 \ -e OAUTH_CLIENT_ID="YOUR_VALUE_HERE" \ -e OAUTH_CLIENT_SECRET="YOUR_VALUE_HERE" \ -e DEFAULT_HF_TOKEN="YOUR_VALUE_HERE" \ -e APP_PORT="5173" \ -e REDIRECT_URI="https://enzostvs-deepsite.hf.space/auth/login" \ registry.hf.space/enzostvs-deepsite:latest

#63 opened about 1 year ago by

huggingface autotrain fail

#61 opened about 1 year ago by

Upload Screenshot_20250404-165732.png

#60 opened about 1 year ago by

Token ID generator tool

#59 opened about 1 year ago by

tendency to hallucinate user input

#57 opened about 1 year ago by

Make config params float to avoid warnings in Transformers

#56 opened about 1 year ago by

تار

#53 opened about 1 year ago by

Downloading weights without duplicates

#52 opened about 1 year ago by

How to run on LM Studio?

#50 opened about 1 year ago by

Move to a park in London

#49 opened about 1 year ago by

Upload 3bd4e13cd467fbed3df8c519a494113c.jpg

#48 opened about 1 year ago by

سلام ازت میخوام این فایل رو بصورت بسیار عمیق و دقیق تحلیل کنی و از تمام زوایا فرمول های اون رو درک کنی و علت هرنوع محاسبه رو بفهمی

#46 opened about 1 year ago by

does the chat template correct for tool use?

#44 opened about 1 year ago by

Copy

#43 opened about 1 year ago by

Amazing, test result by 567Gb RAM on 10 years old hardware

#42 opened about 1 year ago by

Regarding the proposal of "keeping it academic", please close the discussion that does not contain academic/technical information.

#41 opened about 1 year ago by

How to make function calling works?

#39 opened about 1 year ago by

AttributeError: 'DeepseekV3Config' object has no attribute 'seq_aux'

#38 opened about 1 year ago by

听我说谢谢你因为有你温暖了四季

#37 opened about 1 year ago by

Temperature Setting of Different Tasks

#36 opened about 1 year ago by

Upload IMG_2706.jpeg

#34 opened about 1 year ago by

Example for evaluating the DeepSeek-V3-0324 API performance

#33 opened about 1 year ago by