Gguf GGUF models
335 GGUF models in the Gguf family, quantized by 226 uploaders. Sizes run from 1M to 299B parameters. The longest trained context is 1000K tokens.
About the Gguf GGUF models
Gguf is represented on ggufmodels.com by 335 GGUF models, quantized by 226 uploaders on Hugging Face. The most downloaded are gguf pythia 1.4b sft full, gguf components and gguf openhermes tinyllama sft qlora. Parameter counts run from 1M to 299B, and the longest trained context window is 1000K tokens.
A Q4_K_M file of a Gguf model is between 0.0 to 18.6 GB, which is the quant most people start with. Most files come from Felladrin (14), NikolaiRaitschew (6) and Hironabe333 (6). The quants you will meet most often are Q4_K_M (44), Q8_0 (38), F16 (17) and Q5_K_M (16), where the number is how many models offer that quant.
Frequently asked questions
Which Gguf GGUF should I download?
The most downloaded Gguf GGUF here is gguf pythia 1.4b sft full. Pick the size that fits your GPU first, then choose a quant: Q4_K_M is the usual starting point, and Q5_K_M or Q6_K give more quality if you have memory to spare. The VRAM pages below list what fits each GPU.
How much VRAM does a Gguf model need?
Q4_K_M files of Gguf models on this site range from 0.0 to 18.6 GB. Add the KV cache for your context length and about 0.5 GB for the runtime, which usually means 1.5 to 3 GB on top of the file at an 8K context.
Where do Gguf GGUF files come from?
The most common quantizers here are Felladrin, NikolaiRaitschew and Hironabe333. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.