GGUF models by Lewdiculous
162 GGUF models quantized by Lewdiculous on Hugging Face. Sizes run from 3.95B to 14.8B parameters. The longest trained context is 1000K tokens.
About Lewdiculous on Hugging Face
Lewdiculous has published 162 GGUF models that ggufmodels.com tracks, across families such as Llama, Eris and Mistral. The most downloaded are L3-8B-Stheno-v3.2, MN 12B Lyra v4 and Llama 3.1 8B Stheno v3.4.
The quants Lewdiculous publishes most often are Q5_K_M (160), Q4_K_M (160), Q6_K (158) and Q5_K_S (158). Their Q4_K_M files range from 2.5 to 9.0 GB. Parameter counts run from 3.95B to 14.8B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has Lewdiculous published?
162 models are listed here, mainly from the Llama, Eris and Mistral families. The most downloaded is L3-8B-Stheno-v3.2.
Which quants does Lewdiculous publish?
The most common are Q5_K_M (160 models), Q4_K_M (160 models), Q6_K (158 models) and Q5_K_S (158 models).