Quantization publisher
Green-Sky
Green-Sky publishes 9 quantizations across 2 models in our index, averaging 5.177 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 4 of the pairs we can compare — the same label does not mean the same file.
From the file· summed file bytes
Repositories
2
Quantizations
9
Models covered
2
Avg effective bpw
5.177
across their files
Same model, same quant label, different bytes
largest disagreements first
| Model | Quant | Green-Sky | vs | Theirs | Difference |
|---|---|---|---|---|---|
| ERNIE-Image | Q8_0 | 3.80 GiB | ponpoke | 3.40 GiB | +11.7% |
| ERNIE-Image | Q8_0 | 3.80 GiB | ponpoke | 3.40 GiB | +11.7% |
| ERNIE-Image | Q4_K_M | 2.21 GiB | ponpoke | 2.00 GiB | +10.6% |
| ERNIE-Image | Q4_K_M | 2.21 GiB | ponpoke | 2.00 GiB | +10.6% |
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
| Model | Quantizations | Smallest |
|---|---|---|
| ERNIE-Image | 4 | 2.21 GiB |
| SPARK.Chroma_preview | 5 | 4.71 GiB |