GGUF models by Solshine
41 GGUF models quantized by Solshine on Hugging Face. Sizes run from 134M to 13.3B parameters. The longest trained context is 128K tokens.
About Solshine on Hugging Face
Solshine has published 41 GGUF models that ggufmodels.com tracks, across families such as Llama, Jais and InternLM. The most downloaded are Llama 3.2 3B Instruct, Hermes 3 Llama 3.1 8B Lorablated and Llama 3.2 1B.
The quants Solshine publishes most often are Q4_K_M (25), Q8_0 (8), Q2_K (7) and Q4_0 (1). Their Q4_K_M files range from 0.1 to 8.1 GB. Parameter counts run from 134M to 13.3B, and the longest trained context window is 128K tokens.
Frequently asked questions
What GGUF models has Solshine published?
41 models are listed here, mainly from the Llama, Jais and InternLM families. The most downloaded is Llama 3.2 3B Instruct.
Which quants does Solshine publish?
The most common are Q4_K_M (25 models), Q8_0 (8 models), Q2_K (7 models) and Q4_0 (1 models).