Marco GGUF models
121 GGUF models in the Marco family, quantized by 40 uploaders. Sizes run from 4M to 17.4B parameters. The longest trained context is 128K tokens.
About the Marco GGUF models
Marco is represented on ggufmodels.com by 121 GGUF models, quantized by 40 uploaders on Hugging Face. The most downloaded are ms marco MiniLM L6 v2 Q8 0, Marco Mini Instruct and Marco o1. Parameter counts run from 4M to 17.4B, and the longest trained context window is 128K tokens.
A Q4_K_M file of a Marco model is between 0.0 to 10.6 GB, which is the quant most people start with. Most files come from mradermacher (47), sinjab (22) and keisuke-miyako (3). The quants you will meet most often are Q4_K_M (78), Q2_K (66), Q8_0 (64) and Q6_K (64), where the number is how many models offer that quant.
Frequently asked questions
Which Marco GGUF should I download?
The most downloaded Marco GGUF here is ms marco MiniLM L6 v2 Q8 0. 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 Marco model need?
Q4_K_M files of Marco models on this site range from 0.0 to 10.6 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 Marco GGUF files come from?
The most common quantizers here are mradermacher, sinjab and keisuke-miyako. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.