Any Suggestions? any useful info would be Appreciated.
Are you asking about the organisation, the app, the model quants or what?
1 Like
Pretty sure Wendell used their models during the early gpt-OSS days. Haven’t heard much recently
1 Like
For llama.cpp, they’re basically the default source for high-quality GGUFs as far as I’m concerned.
Hell, even now that I’m using vllm-radiance, I’m their FP8 quant of Qwen 3.8 27B.
4 Likes
Their releases would be more respectable if they measured and published KLD numbers so people downloading their quants could tell how degraded they are wrt the original model.
3 Likes
No just tips to use it.