GGUF models by hardlyworking
37 GGUF models quantized by hardlyworking on Hugging Face. Sizes run from 4.02B to 32.8B parameters. The longest trained context is 1000K tokens.
About hardlyworking on Hugging Face
hardlyworking has published 37 GGUF models that ggufmodels.com tracks, across families such as Final, Ramen and Holdmy. The most downloaded are Qwen3 32B Biprojected Norm Preserving Abliterated, 4Brp Q4 0 and Secret4B.
The quants hardlyworking publishes most often are Q4_0 (14), Q4_K_S (7), Q8_0 (6) and Q4_K_M (4). Their Q4_K_M files range from 14.3 to 19.8 GB. Parameter counts run from 4.02B to 32.8B, and the longest trained context window is 1000K tokens.
Frequently asked questions
What GGUF models has hardlyworking published?
37 models are listed here, mainly from the Final, Ramen and Holdmy families. The most downloaded is Qwen3 32B Biprojected Norm Preserving Abliterated.
Which quants does hardlyworking publish?
The most common are Q4_0 (14 models), Q4_K_S (7 models), Q8_0 (6 models) and Q4_K_M (4 models).