Quantization publisher

lordx64

lordx64 publishes 14 quantizations across 4 models in our index, averaging 5.795 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 9 of the pairs we can compare — the same label does not mean the same file.

From the file· summed file bytes
Repositories
6
Quantizations
14
Models covered
4
Avg effective bpw
5.795
across their files

Same model, same quant label, different bytes

largest disagreements first
ModelQuantlordx64vsTheirsDifference
Qwable-v1Q8_034.37 GiBbartowski35.22 GiB-2.4%
Qwable-v2Q8_034.37 GiBmradermacher35.21 GiB-2.4%
Qwen3.6-35B-A3B-Claude-4.7-Opus-Reasoning-DistilledQ8_034.37 GiBhuihui-ai35.21 GiB-2.4%
Qwable-v1Q5_K_M23.03 GiBbartowski23.74 GiB-3.0%
Qwable-v1Q4_K_M19.71 GiBbartowski20.36 GiB-3.2%
Qwable-v2Q5_K_M23.03 GiBmradermacher23.61 GiB-2.4%
Qwable-v2Q4_K_M19.71 GiBmradermacher20.22 GiB-2.5%
Qwable-v2IQ4_XS17.64 GiBmradermacher18.06 GiB-2.3%
Qwable-v1IQ4_XS17.64 GiBbartowski17.95 GiB-1.8%

A quantization label describes a target, not a recipe. Publishers make different choices about which tensors to keep at higher precision, and some apply an importance matrix while others don't — so two files both honestly labelled the same thing can differ measurably in size and in quality.

Models they publish