Starling GGUF models
71 GGUF models in the Starling family, quantized by 33 uploaders. Sizes run from 6.74B to 46.7B parameters. The longest trained context is 128K tokens.
About the Starling GGUF models
Starling is represented on ggufmodels.com by 71 GGUF models, quantized by 33 uploaders on Hugging Face. The most downloaded are berkeley-nest_starling-lm-7b-alpha, Starling-LM-7B-alpha-x2 and Starling LM 7B alpha ExPO. Parameter counts run from 6.74B to 46.7B, and the longest trained context window is 128K tokens.
A Q4_K_M file of a Starling model is between 4.1 to 28.6 GB, which is the quant most people start with. Most files come from mradermacher (25), RichardErkhov (6) and tensorblock (4). The quants you will meet most often are Q4_K_M (58), Q2_K (53), Q5_K_M (52) and Q6_K (51), where the number is how many models offer that quant.
Frequently asked questions
Which Starling GGUF should I download?
The most downloaded Starling GGUF here is berkeley-nest_starling-lm-7b-alpha. 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 Starling model need?
Q4_K_M files of Starling models on this site range from 4.1 to 28.6 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 Starling GGUF files come from?
The most common quantizers here are mradermacher, RichardErkhov and tensorblock. They convert the original model and publish quantized files on Hugging Face. ggufmodels.com does not host files; every download button opens Hugging Face.