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
dalatexcoder's picture
🔄 In a p-e-w/heretic and mlx-optiq loop

dalatexcoder

dalatexcoder
2 1
rodoviario's profile picture eac123's profile picture lyraaaa's profile picture
·
  • Dalatexcoder

AI & ML interests

Small model (under 3B) abliteration (via heretic-llm with custom parameters) and quantization.

Recent Activity

updated a collection 19 days ago
Research Models
updated a collection 19 days ago
Research Models
updated a collection 19 days ago
Research Models
View all activity

Organizations

MLX Community's profile picture

updated 4 collections 19 days ago

Research Models

Collection
Not production grade. No guarantees on quality. May not even spit out coherent speech. • 5 items • Updated 19 days ago

Quantized, Ready to use (MLX)

Collection
Quantized (using optiq's mixed bit calibration mostly). Base models are not always created by me. • 8 items • Updated 19 days ago

Abliterated, Quantized Production Grade Models

Collection
Decensored, Quantized, ready to use. MLX and abliteration by me, GGUFs by mradermacher. Mixed perc preferred. Usable models. • 6 items • Updated 19 days ago

Abliterated, Production Grade Models

Collection
These models are abliterated via Heretic and are uncensored while preserving capability for ordinary tasks. Please use at your own risk. BF16 Versions • 7 items • Updated 19 days ago
Company
TOS Privacy About Careers
Website
Models Datasets Spaces Pricing Docs