Pythia GGUF models
303 GGUF models in the Pythia family, quantized by 26 uploaders. Sizes run from 14M to 11.8B parameters. The longest trained context is 9K tokens.
About the Pythia GGUF models
Pythia is represented on ggufmodels.com by 303 GGUF models, quantized by 26 uploaders on Hugging Face. The most downloaded are Pythia 6.9b, Pythia 12b Deduped and Pythia 410m. Parameter counts run from 14M to 11.8B, and the longest trained context window is 9K tokens.
A Q4_K_M file of a Pythia model is between 0.0 to 7.6 GB, which is the quant most people start with. Most files come from mradermacher (130), RichardErkhov (86) and tensorblock (41). The quants you will meet most often are Q2_K (264), Q4_K_M (235), Q3_K_M (235) and Q4_K_S (226), where the number is how many models offer that quant.
Frequently asked questions
Which Pythia GGUF should I download?
The most downloaded Pythia GGUF here is Pythia 6.9b. 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 Pythia model need?
Q4_K_M files of Pythia models on this site range from 0.0 to 7.6 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 Pythia 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.