[johndough]

> it gets caught in circles more often then say a 1t parameter model does.

I've found that the Q5+ quants are less loopy than Q4. Still not perfect, but noticeably better.

> reasonable enough tokens per second

The speed has been amazing. I've been running the recent llama.cpp MTP branch with an uncensored variant of Qwen3.6-35B-A3B on my RTX 3090 over 170 tokens per second and it was able to turn a buffer overflow into a reliable shell exploit in just a few seconds (with reasoning disabled). Still a bit loopy though. Hopefully, the Qwen team will pay more attention to those looping issues. It feels like their models are especially susceptible.

[2ndorderthought]

Is that on a single 3090? I need to change my settings it sounds like

[johndough]

Yes, single RTX 3090 with this model https://huggingface.co/llmfan46/Qwen3.6-35B-A3B-uncensored-h... following these https://huggingface.co/havenoammo/Qwen3.6-35B-A3B-MTP-GGUF instructions (should add "-j 8" to last cmake command for parallel build) and llama-server with --reasoning off

Note that the MTP PR https://github.com/ggml-org/llama.cpp/pull/22673 is still under development, so things might be broken.