Quantization publisher

Serveurperso

Serveurperso publishes 13 quantizations across 3 models in our index, averaging 13.970 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 1 of the pairs we can compare — the same label does not mean the same file.

From the file· summed file bytes
Repositories
3
Quantizations
13
Models covered
3
Avg effective bpw
13.970
across their files

Same model, same quant label, different bytes

largest disagreements first
ModelQuantServeurpersovsTheirsDifference
OmniVoiceQ8_00.88 GiBcstr1.06 GiB-16.9%

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

ModelQuantizationsSmallest
Ace-Step1.5519.22 GiB
Qwen3-TTS-12Hz-0.6B-Base44.72 GiB
OmniVoice40.61 GiB