> Second, besides noise (bars are Wilson 95% confidence intervals, very conservative for run-to-run noise), there is little difference down to 4-bit; only the 2-bit scores a bit lower.

Confidence intervals have nothing to do with run-to-run variation. They have little to do with anything people usually ascribe to them (https://link.springer.com/article/10.3758/s13423-015-0947-8 ), but even less with run-to-run variation (https://link.springer.com/article/10.1007/s10654-016-0149-3 misconception 22).

This confirms a theory I have to explain the minimal loss in quality when using lower quants (I use IQ3_XXS with an 8-bit KV cache) and the XHIGH (default) thinking level.

It's well-known that while quantization affects the sampling probability distribution (given the same context, which next token is the most probable), Qwen 3.8 27b seems to offset that by just thinking more and as a result eventually finishing the task (benchmark or otherwise).

So as long as the thinking (albeit longer) is sound, this leads to the same success rate (as shown in the article) but potentially at the cost of more tokens and hence more time.

I think it'll be further useful to chart each quantization's used tokens as well, in addition to the success rate.

Thanks for doing and sharing the research!

There's a real hole here at Q3. A critical breakpoint here is sub 16-GB cards, which covers the 5080, 5070 Ti, 5060ti, and several other cards from this generation and the last. It would be instructive to see where the quality knee is.

hmm, assuming that this article is part written by claude and part human-written, can anyone help me find a rule of thumb for "how to know if the article is worth reading"?

Because on the one hand, the prose and the presentation is painful (narrating irrelevant points, nonlinear X-axes, ambiguous chart labels, etc etc),

But on the other hand, the result that I'm assuming the author means to communicate ("on these evals, generation quality seems fairly good") sounds worthwhile to share?

Because I really struggle with this question at the moment. Am I allowed to draw an adverse inference that "if the writeup presents irrelevant text side by side with the data, then this may be a sign that the author does not understand the task that they are attempting to write up"?

Note that these quants are not quantized uniformly, so 4-bit isn't actually a "true" 4-bit here, so these observations won't necessarily hold up to other quants which might be done differently.

I’m more curious how each 4 bit quant compares. It seems like NVFP4 outperforms Q4_K_M in terms of speed and top 1 but is only good for expensive Nvidia cards

Those benchmarks are very interesting.

But is there any model that actually works in a decent way at quantization of 1?

Could this be used to estimate how many fingers LLM have?

Qwen3.8 27B seems like it was clearly supposed to be a high-end consumer open-weights model, but the t/s is so low for me on my old M1 Max 64GB that I hope others are getting use out of it.

Unfortunately, the calculus has changed and it seems cheaper to me to just use MiMo V2.5 for pennies or DeepSeek V4 Flash instead of using Qwen anymore unless I need a local model specifically for doing reverse engineering work that gets otherwise rejected.

the 4-bit matching bf16 on terminal-bench is a useful data

I've been running Unsloth IQ3_S on my 5060ti with mmproj offloaded, getting 600-1000 prefill and 30-50 tg with this config:

       /data/llm/llama.cpp/build/bin/llama-server
        --threads 4
        --threads-batch 8
        --batch-size 4096
        --ubatch-size 256
        --port 9999
        --temp "1.0"
        --top-p "0.95"
        --top-k "20"
        --min-p "0.0"
        --presence-penalty "0.0"
        --reasoning auto
        --reasoning-preserve
        --reasoning-budget 4096
        --gpu-layers-draft all
        --spec-type draft-mtp,ngram-map-k4v,ngram-mod
        --spec-draft-n-max 3
        --spec-draft-p-min 0.75
        --spec-ngram-mod-n-match 24
        --spec-ngram-mod-n-min 4
        --spec-ngram-mod-n-max 16
        --spec-ngram-map-k4v-size-n 8
        --spec-ngram-map-k4v-size-m 16
        --spec-ngram-map-k4v-min-hits 1
        --n-gpu-layers all
        --ctx-size 131072
        --repeat-penalty 1.0
        --jinja
        --metrics
        --model /data/llm/models/unsloth/Qwen3.8-27B-UD-IQ3_S.gguf
        --chat-template-file /data/llm/models/qwen3.6-chat-template.jinja
        --fit off
        --flash-attn on
        --cors-origins localhost
        --mmproj /data/llm/models/unsloth/Qwen3.8/mmproj-BF16.gguf
        --no-mmproj-offload
        --parallel 1
        --kv-unified
        --cache-type-k q4_0
        --cache-type-v q4_0
        --cache-type-k-draft q4_0
        --cache-type-v-draft q4_0

local llm don't make sense currently consumer compute is not upto mark it may take atleast 7 more years to be usable