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
      • Hardware
      • Learn
      • Discord
      • Forum
      • GitHub
    • Solutions
      • Team & Enterprise
      • Hugging Face PRO
      • Enterprise Support
      • Inference Providers
      • Inference Endpoints
      • Storage Buckets

  • Log In
  • Sign Up

EximiusLabs
/
fusion-embedding-2-2b-preview

Feature Extraction
Safetensors
English
fusion-embedding-connector
embeddings
multimodal
audio
retrieval
matryoshka
qwen3-vl
adapters
custom_code
Model card Files Files and versions
xet
Community
fusion-embedding-2-2b-preview
485 MB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 41 commits
abtonmoy's picture
abtonmoy
v0.3-preview: retrieval-tuned flagship
1720d8b verified about 2 hours ago
  • assets
    Positioning figure: trained-parameters axis (matches the technical report) 12 days ago
  • .gitattributes
    1.7 kB
    assets: unified-model positioning chart 12 days ago
  • README.md
    9.4 kB
    v0.3-preview: retrieval-tuned flagship about 2 hours ago
  • config.json
    1.96 kB
    v0.3-preview: retrieval-tuned flagship about 2 hours ago
  • configuration_fusion_embedding.py
    2.49 kB
    AutoModel + trust_remote_code support (remote-code loading; existing loaders unchanged) 12 days ago
  • fusion-embedding-2-2b-preview.pt
    242 MB
    xet
    v0.3-preview: retrieval-tuned flagship about 2 hours ago
  • inference.py
    19 kB
    inference.py: native video preprocessing (no qwen-vl-utils) 10 days ago
  • model.safetensors
    242 MB
    xet
    v0.3-preview: retrieval-tuned flagship about 2 hours ago
  • modeling_fusion_embedding.py
    33.2 kB
    embed_video: native preprocessing (frame tensors and torchcodec path decode); drops the qwen-vl-utils dependency; bitwise-equal to the previous implementation and the raw base on identical inputs 10 days ago