Model comparison

stable-diffusion-v1-5 vs Chroma1-HD

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

From the file· summed bytes, KV per layer

Side by side

stable-diffusion-v1-5Chroma1-HD
Parameters860M8.9B
Architectureflux
Layers
Native context
Mixture of expertsnono
Quantizations published718
Smallest quantization1.46 GiB3.18 GiB
Q4_K_M5.18 GiB
Licencecreativeml-openrail-mapache-2.0

KV cache by context

the term that decides long-context viability
Contextstable-diffusion-v1-5Chroma1-HDRatio
4,096
8,192
16,384
32,768
65,536
131,072