Model comparison
Anima vs stable-diffusion-v1-5
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| Anima | stable-diffusion-v1-5 | |
|---|---|---|
| Parameters | 2.1B | 860M |
| Architecture | cosmos | — |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 23 | 7 |
| Smallest quantization | 0.95 GiB | 1.46 GiB |
| Q4_K_M | 1.29 GiB | — |
| Licence | other | creativeml-openrail-m |
KV cache by context
the term that decides long-context viability
| Context | Anima | stable-diffusion-v1-5 | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |