GGUF models by grimjim
38 GGUF models quantized by grimjim on Hugging Face. Sizes run from 7.24B to 23.6B parameters. The longest trained context is 1000K tokens.
About grimjim on Hugging Face
grimjim has published 38 GGUF models that ggufmodels.com tracks, across families such as Llama, Kukulemon and Gemma. The most downloaded are Llama 3.1 8B Instruct Abliterated_Via_Adapter, Magnolia Mell v1 12B and cuckoo starling 32k 7B.
The quants grimjim publishes most often are Q8_0 (37), Q6_K (34), Q5_K_M (34) and Q4_K_M (34). Their Q4_K_M files range from 4.4 to 14.3 GB. Parameter counts run from 7.24B to 23.6B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has grimjim published?
38 models are listed here, mainly from the Llama, Kukulemon and Gemma families. The most downloaded is Llama 3.1 8B Instruct Abliterated_Via_Adapter.
Which quants does grimjim publish?
The most common are Q8_0 (37 models), Q6_K (34 models), Q5_K_M (34 models) and Q4_K_M (34 models).