Quantization publisher

Yingyaeliae

Yingyaeliae publishes 10 quantizations across 2 models in our index, averaging 5.537 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 8 of the pairs we can compare — the same label does not mean the same file.

From the file· summed file bytes
Repositories
2
Quantizations
10
Models covered
2
Avg effective bpw
5.537
across their files

Same model, same quant label, different bytes

largest disagreements first
ModelQuantYingyaeliaevsTheirsDifference
grok-oss-Apollyon-24BQ8_023.33 GiBmradermacher23.33 GiB-0.0%
grok-oss-Apollyon-24BQ3_K_M10.69 GiBmradermacher10.69 GiB-0.0%
grok-oss-Apollyon-24BQ4_K_M13.35 GiBmradermacher13.35 GiB-0.0%
grok-oss-Apollyon-24BQ5_K_M15.61 GiBmradermacher15.61 GiB-0.0%
grok-oss-Apollyon-8BQ3_K_M3.74 GiBmradermacher3.74 GiB-0.0%
grok-oss-Apollyon-8BQ4_K_M4.58 GiBmradermacher4.58 GiB-0.0%
grok-oss-Apollyon-8BQ5_K_M5.34 GiBmradermacher5.34 GiB-0.0%
grok-oss-Apollyon-8BQ8_07.95 GiBmradermacher7.95 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
grok-oss-Apollyon-8B53.74 GiB
grok-oss-Apollyon-24B510.69 GiB