Model comparison

SmolLM-135M-Instruct vs ced-base

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

SmolLM-135M-Instructced-base
Parameters135M86M
Architecturellamaced
Layers30
Native context2,048
Mixture of expertsnono
Quantizations published243
Smallest quantization0.08 GiB0.12 GiB
Q4_K_M0.10 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextSmolLM-135M-Instructced-baseRatio
4,0960.09 GiB
8,1920.18 GiB
16,3840.35 GiB
32,7680.70 GiB
65,5361.41 GiB
131,0722.81 GiB