Quantization publisher

8sp4rk

8sp4rk publishes 5 quantizations across 1 models in our index, averaging 5.745 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 5 of the pairs we can compare — the same label does not mean the same file.

From the file· summed file bytes
Repositories
1
Quantizations
5
Models covered
1
Avg effective bpw
5.745
across their files

Same model, same quant label, different bytes

largest disagreements first
ModelQuant8sp4rkvsTheirsDifference
Qwopus3.6-27B-CoderQ8_026.63 GiBJackrong27.05 GiB-1.6%
Qwopus3.6-27B-CoderQ6_K20.57 GiBJackrong20.89 GiB-1.6%
Qwopus3.6-27B-CoderQ5_K_M17.91 GiBJackrong18.19 GiB-1.6%
Qwopus3.6-27B-CoderQ4_K_M15.41 GiBJackrong15.66 GiB-1.6%
Qwopus3.6-27B-CoderQ3_K_M12.39 GiBJackrong12.57 GiB-1.5%

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
Qwopus3.6-27B-Coder512.39 GiB