GGUF models by DevQuasar-4
48 GGUF models quantized by DevQuasar-4 on Hugging Face. Sizes run from 494M to 426B parameters. The longest trained context is 1024K tokens.
About DevQuasar-4 on Hugging Face
DevQuasar-4 has published 48 GGUF models that ggufmodels.com tracks, across families such as Llama, Qwen and Mistral. The most downloaded are Huihui ai.DeepSeek R1 Distill Qwen 32B Abliterated, Huihui ai.Qwen2.5 7B Instruct 1M Abliterated and Huihui ai.DeepSeek V3 0324 Pruned Coder 411B.
The quants DevQuasar-4 publishes most often are Q5_K_M (48), Q4_K_M (48), Q3_K_M (48) and Q8_0 (47). Their Q4_K_M files range from 0.3 to 256.6 GB. Parameter counts run from 494M to 426B, and the longest trained context window is 1024K tokens.
Frequently asked questions
What GGUF models has DevQuasar-4 published?
48 models are listed here, mainly from the Llama, Qwen and Mistral families. The most downloaded is Huihui ai.DeepSeek R1 Distill Qwen 32B Abliterated.
Which quants does DevQuasar-4 publish?
The most common are Q5_K_M (48 models), Q4_K_M (48 models), Q3_K_M (48 models) and Q8_0 (47 models).