GGUF models by maddes8cht
88 GGUF models quantized by maddes8cht on Hugging Face. Sizes run from 1.59B to 41.9B parameters. The longest trained context is 64K tokens.
About maddes8cht on Hugging Face
maddes8cht has published 88 GGUF models that ggufmodels.com tracks, across families such as Falcon, Llama and Mpt. The most downloaded are Falcon, acrastt Puma 3B and Falcon.
The quants maddes8cht publishes most often are Q8_0 (86), Q5_K_M (86), Q4_K_M (86) and Q4_0 (86). Their Q4_K_M files range from 1.0 to 25.5 GB. Parameter counts run from 1.59B to 41.9B, and the longest trained context window is 64K tokens.
Frequently asked questions
What GGUF models has maddes8cht published?
88 models are listed here, mainly from the Falcon, Llama and Mpt families. The most downloaded is Falcon.
Which quants does maddes8cht publish?
The most common are Q8_0 (86 models), Q5_K_M (86 models), Q4_K_M (86 models) and Q4_0 (86 models).