Experiment GGUF models
393 GGUF models in the Experiment family, quantized by 20 uploaders. Sizes run from 3.09B to 70.6B parameters. The longest trained context is 256K tokens.
About the Experiment GGUF models
Experiment is represented on ggufmodels.com by 393 GGUF models, quantized by 20 uploaders on Hugging Face. The most downloaded are Experiment26 7B passthrough 10slice, models--automerger--Experiment28Experiment29-7B and Experiment26 7B. Parameter counts run from 3.09B to 70.6B, and the longest trained context window is 256K tokens.
A Q4_K_M file of a Experiment model is between 2.0 to 42.5 GB, which is the quant most people start with. Most files come from mradermacher (303), MaziyarPanahi (43) and RichardErkhov (15). The quants you will meet most often are Q4_K_M (380), Q2_K (375), Q5_K_M (374) and Q3_K_M (372), where the number is how many models offer that quant.
Frequently asked questions
Which Experiment GGUF should I download?
The most downloaded Experiment GGUF here is Experiment26 7B passthrough 10slice. 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 Experiment model need?
Q4_K_M files of Experiment models on this site range from 2.0 to 42.5 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 Experiment GGUF files come from?
The most common quantizers here are mradermacher, MaziyarPanahi and RichardErkhov. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.