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

  • Log In
  • Sign Up

abstract-extraordinary
/
Qwen3.6-27B-DSpark

Transformers
Safetensors
qwen3
speculative-decoding
dspark
draft-model
vllm
text-generation-inference
Model card Files Files and versions
xet
Community

Instructions to use abstract-extraordinary/Qwen3.6-27B-DSpark with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

  • Libraries
  • Transformers

    How to use abstract-extraordinary/Qwen3.6-27B-DSpark with Transformers:

    # pip install -U transformers accelerate
    # Load model directly
    from transformers import AutoTokenizer, AutoModel
    
    tokenizer = AutoTokenizer.from_pretrained("abstract-extraordinary/Qwen3.6-27B-DSpark")
    model = AutoModel.from_pretrained("abstract-extraordinary/Qwen3.6-27B-DSpark", device_map="auto")
  • Notebooks
  • Google Colab
  • Kaggle
Qwen3.6-27B-DSpark
8.8 GB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 2 commits
stepnivlk's picture
stepnivlk
DSpark v2 head, full vocab: 2.62x vs no drafter, 4.67 tok/step
cb46d3b verified 3 months ago
  • .gitattributes
    1.52 kB
    initial commit 3 months ago
  • README.md
    8.19 kB
    DSpark v2 head, full vocab: 2.62x vs no drafter, 4.67 tok/step 3 months ago
  • acceptance-by-position.png
    68.4 kB
    DSpark v2 head, full vocab: 2.62x vs no drafter, 4.67 tok/step 3 months ago
  • config.json
    1.62 kB
    DSpark v2 head, full vocab: 2.62x vs no drafter, 4.67 tok/step 3 months ago
  • model.safetensors
    8.8 GB
    xet
    DSpark v2 head, full vocab: 2.62x vs no drafter, 4.67 tok/step 3 months ago