Quantization publisher

mlabonne

mlabonne publishes 6 quantizations across 1 models in our index, averaging 5.518 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 7 of the pairs we can compare — the same label does not mean the same file.

From the file· summed file bytes
Repositories
1
Quantizations
6
Models covered
1
Avg effective bpw
5.518
across their files

Same model, same quant label, different bytes

largest disagreements first
ModelQuantmlabonnevsTheirsDifference
gemma-3-12b-it-abliterated-v2Q4_K_M6.80 GiBmvyzjoph6.80 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q3_K_M5.60 GiBDevQuasar5.60 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q4_K_M6.80 GiBDevQuasar6.80 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q2_K4.44 GiBDevQuasar4.44 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q5_K_M7.87 GiBDevQuasar7.87 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q6_K9.00 GiBDevQuasar9.00 GiB-0.0%
gemma-3-12b-it-abliterated-v2Q8_011.65 GiBDevQuasar11.65 GiB-0.0%

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
gemma-3-12b-it-abliterated-v264.44 GiB