GGUF models by djuna
41 GGUF models quantized by djuna on Hugging Face. Sizes run from 32M to 27.2B parameters. The longest trained context is 1000K tokens.
About djuna on Hugging Face
djuna has published 41 GGUF models that ggufmodels.com tracks, across families such as Test, Gemma and Qwen. The most downloaded are Artigenz Coder DS 6.7B Q5 K M, jina embeddings v2 base en Q5 K M and MN Chinofun 12B 4.1 Q6 K.
The quants djuna publishes most often are Q5_K_M (26), Q8_0 (3), Q6_K (3) and Q4_K_S (3). Their Q4_K_M files range from 4.7 to 7.5 GB. Parameter counts run from 32M to 27.2B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has djuna published?
41 models are listed here, mainly from the Test, Gemma and Qwen families. The most downloaded is Artigenz Coder DS 6.7B Q5 K M.
Which quants does djuna publish?
The most common are Q5_K_M (26 models), Q8_0 (3 models), Q6_K (3 models) and Q4_K_S (3 models).