Hugging Face's logo Hugging Face
  • Models
  • Datasets
  • Spaces
  • Buckets new
  • Docs
  • Enterprise
  • Pricing
    • Website
      • Tasks
      • HuggingChat
      • Collections
      • Languages
      • Organizations
    • Community
      • Blog
      • Posts
      • Daily Papers
      • Learn
      • Discord
      • Forum
      • GitHub
    • Solutions
      • Team & Enterprise
      • Hugging Face PRO
      • Enterprise Support
      • Inference Providers
      • Inference Endpoints
      • Storage Buckets

  • Log In
  • Sign Up

SGLang
/
DeepSeek-V3-NextN

Transformers
Safetensors
deepseek_v3
custom_code
text-generation-inference
fp8
Model card Files Files and versions
xet
Community
4

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

  • Libraries
  • Transformers

    How to use SGLang/DeepSeek-V3-NextN with Transformers:

    # Load model directly
    from transformers import AutoTokenizer, DeepseekV3ForCausalLMNextN
    
    tokenizer = AutoTokenizer.from_pretrained("SGLang/DeepSeek-V3-NextN", trust_remote_code=True)
    model = DeepseekV3ForCausalLMNextN.from_pretrained("SGLang/DeepSeek-V3-NextN", trust_remote_code=True)
  • Notebooks
  • Google Colab
  • Kaggle
New discussion
Resources
  • PR & discussions documentation
  • Code of Conduct
  • Hub documentation

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

#4 opened 9 days ago by
qgallouedec

Has anyone achieved a speed-up with this model?

#3 opened about 1 year ago by
RonanMcGovern

Add text-generation pipeline tag and MIT license

#2 opened about 1 year ago by
nielsr

Is this MTP head just for predicting one token ahead?

#1 opened about 1 year ago by
RonanMcGovern
Company
TOS Privacy About Careers
Website
Models Datasets Spaces Pricing Docs