GGUF models by Stark2008
48 GGUF models quantized by Stark2008 on Hugging Face. Sizes run from 3.82B to 21.4B parameters. The longest trained context is 1000K tokens.
About Stark2008 on Hugging Face
Stark2008 has published 48 GGUF models that ggufmodels.com tracks, across families such as Llama, Dictalm and Mistral. The most downloaded are Phi 3 Medium 4k Instruct Abliterated v3, Chocolatine 3B Instruct DPO Revised Q8 0 and Qwen1.5-14B-Chat.
The quants Stark2008 publishes most often are Q6_K (29), Q3_K_L (4), Q4_K_M (3) and Q8_0 (2). Their Q4_K_M files range from 4.4 to 9.2 GB. Parameter counts run from 3.82B to 21.4B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has Stark2008 published?
48 models are listed here, mainly from the Llama, Dictalm and Mistral families. The most downloaded is Phi 3 Medium 4k Instruct Abliterated v3.
Which quants does Stark2008 publish?
The most common are Q6_K (29 models), Q3_K_L (4 models), Q4_K_M (3 models) and Q8_0 (2 models).