GGUF models by PJRM
60 GGUF models quantized by PJRM on Hugging Face. Sizes run from 1.54B to 4.51B parameters. The longest trained context is 256K tokens.
About PJRM on Hugging Face
PJRM has published 60 GGUF models that ggufmodels.com tracks, across families such as Llama, Qwen and Stablelm. The most downloaded are Qwen3.8 4B Distill Heretic, Llama 3.2 3B Instruct and MiniMax M2.5 Tiny 24e.
The quants PJRM publishes most often are Q4_0 (36), IQ4_NL (15), Q4_K_M (10) and Q6_K (1). Their Q4_K_M files range from 1.0 to 2.8 GB. Parameter counts run from 1.54B to 4.51B, and the longest trained context window is 256K tokens.
Frequently asked questions
What GGUF models has PJRM published?
60 models are listed here, mainly from the Llama, Qwen and Stablelm families. The most downloaded is Qwen3.8 4B Distill Heretic.
Which quants does PJRM publish?
The most common are Q4_0 (36 models), IQ4_NL (15 models), Q4_K_M (10 models) and Q6_K (1 models).