Quantization publisher
HuggingFaceTB
HuggingFaceTB publishes 4 quantizations across 4 models in our index, averaging 7.659 effective bits per weight. Their files differ in size from other publishers' builds of the same nominal quantization on 6 of the pairs we can compare — the same label does not mean the same file.
From the file· summed file bytes
Repositories
4
Quantizations
4
Models covered
4
Avg effective bpw
7.659
across their files
Same model, same quant label, different bytes
largest disagreements first
| Model | Quant | HuggingFaceTB | vs | Theirs | Difference |
|---|---|---|---|---|---|
| SmolLM-135M-Instruct | Q8_0 | 0.13 GiB | second-state | 0.13 GiB | 0.0% |
| SmolLM-135M-Instruct | Q8_0 | 0.13 GiB | MaziyarPanahi | 0.13 GiB | 0.0% |
| SmolLM2-1.7B-Instruct | Q4_K_M | 0.98 GiB | bartowski | 0.98 GiB | -0.0% |
| SmolLM2-360M-Instruct | Q8_0 | 0.36 GiB | bartowski | 0.36 GiB | -0.0% |
| smollm-360M-instruct-add-basics | Q8_0 | 0.36 GiB | Felladrin | 0.36 GiB | -0.0% |
| SmolLM2-360M-Instruct | Q8_0 | 0.36 GiB | unsloth | 0.36 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
| Model | Quantizations | Smallest |
|---|---|---|
| SmolLM2-1.7B-Instruct | 1 | 0.98 GiB |
| SmolLM2-360M-Instruct | 1 | 0.36 GiB |
| SmolLM-135M-Instruct | 1 | 0.13 GiB |
| smollm-360M-instruct-add-basics | 1 | 0.36 GiB |