GGUF models by SinpxAI
50 GGUF models quantized by SinpxAI on Hugging Face. Sizes run from 1.1B to 7.51B parameters. The longest trained context is 195K tokens.
About SinpxAI on Hugging Face
SinpxAI has published 50 GGUF models that ggufmodels.com tracks, across families such as Mistral, Tinyllama and Llama. The most downloaded are Llama2 7B Chat uncensored, CodeLlama 7B Python and Dolphin Mistral 7B v2.2.1.
The quants SinpxAI publishes most often are Q8_0 (50), Q6_K (50), Q5_K_S (50) and Q5_K_M (50). Their Q4_K_M files range from 0.7 to 4.6 GB. Parameter counts run from 1.1B to 7.51B, and the longest trained context window is 195K tokens.
Frequently asked questions
What GGUF models has SinpxAI published?
50 models are listed here, mainly from the Mistral, Tinyllama and Llama families. The most downloaded is Llama2 7B Chat uncensored.
Which quants does SinpxAI publish?
The most common are Q8_0 (50 models), Q6_K (50 models), Q5_K_S (50 models) and Q5_K_M (50 models).