3 ms·
I believe you mean 35B-A3B, there was no such thing as A4B. I use 27B and other models for software development, and quite a few research or similar agents. I c
by smcleod 1mo ago
I believe you mean 35B-A3B, there was no such thing as A4B. I use 27B and other models for software development, and quite a few research or similar agents. I cannot imagine a world where the old 35B-A3B model is smarter / more capable than 3.8 27B - the difference is night and day for coding at least. Where 35B-A3B was fast and felt like a Haiku model, 27B feels like a strong Sonnet when given the right tools. I don't use turbo quant so can't comment on that, but with the A3B model you're using you probably won't get much from using MTP with small MoE models like that.
- xscott 1mo agoPeople over-quantize things, muck with the temperature and other settings based on superstitions or results from models they think are similar. There's lots of ways to make 3.8 27B dumber.
- razster 1mo agoI've had great success with allowing the model to review optimal settings and my system specs. It comes up with the right configuration. Running Pi harness. I just had Ornith 1.5 take a moment to configure itself, now it's reviewing a large project I'm working on, so far its really impressive for my needs. Qwen3.8 27b Unsloth(Dynamic 3.0) is also perfect. These two are working together, and I'm in a sweet-spot, I now have all I need.
- seanmcdirmid 1mo agoYes, A3B, sorry for the typo. I’ve benched it and used it in practice for agentic coding, and I haven’t seen any benefit to it yet. Lightning MTP helps a lot in the perf department.