GGUF models by shidowake
36 GGUF models quantized by shidowake on Hugging Face. Sizes run from 2.8B to 13.1B parameters. The longest trained context is 32K tokens.
About shidowake on Hugging Face
shidowake has published 36 GGUF models that ggufmodels.com tracks, across families such as Test, Swal and Mistral. The most downloaded are 240306 swal 13b bnb capybara 6k gpt4 alpaca 6k qlora adaptor fp16merged, japanese stablelm base gamma 7b base and gemma 7b base.
The quants shidowake publishes most often are Q8_0 (21), Q4_K_M (10), Q6_K (8) and Q5_K_M (8). Their Q4_K_M files range from 4.1 to 7.9 GB. Parameter counts run from 2.8B to 13.1B, and the longest trained context window is 32K tokens.
Frequently asked questions
What GGUF models has shidowake published?
36 models are listed here, mainly from the Test, Swal and Mistral families. The most downloaded is 240306 swal 13b bnb capybara 6k gpt4 alpaca 6k qlora adaptor fp16merged.
Which quants does shidowake publish?
The most common are Q8_0 (21 models), Q4_K_M (10 models), Q6_K (8 models) and Q5_K_M (8 models).