GGUF models by Nekuromento
75 GGUF models quantized by Nekuromento on Hugging Face. Sizes run from 494M to 16B parameters. The longest trained context is 128K tokens.
About Nekuromento on Hugging Face
Nekuromento has published 75 GGUF models that ggufmodels.com tracks, across families such as Hammer, Llama and Ghost. The most downloaded are Llama 3.1 WhiteRabbitNeo 2 8B, Codefuse Starcoder2 15b and Athene Codegemma 2 7b It Alpaca v1.
The quants Nekuromento publishes most often are Q6_K (23), Q5_K_M (20), Q8_0 (16) and Q4_K_M (16). Their Q4_K_M files range from 1.9 to 9.9 GB. Parameter counts run from 494M to 16B, and the longest trained context window is 128K tokens.
Frequently asked questions
What GGUF models has Nekuromento published?
75 models are listed here, mainly from the Hammer, Llama and Ghost families. The most downloaded is Llama 3.1 WhiteRabbitNeo 2 8B.
Which quants does Nekuromento publish?
The most common are Q6_K (23 models), Q5_K_M (20 models), Q8_0 (16 models) and Q4_K_M (16 models).