Model comparison
moonshine-streaming-tiny vs moonshine-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
| moonshine-streaming-tiny | moonshine-base | |
|---|---|---|
| Parameters | 44M | 62M |
| Architecture | moonshine_streaming | moonshine |
| Layers | 6 | — |
| Native context | 4,096 | 194 |
| Mixture of experts | no | no |
| Quantizations published | 4 | 5 |
| Smallest quantization | 0.03 GiB | 0.04 GiB |
| Q4_K_M | — | — |
| Licence | mit | mit |
KV cache by context
the term that decides long-context viability
| Context | moonshine-streaming-tiny | moonshine-base | Ratio |
|---|---|---|---|
| 4,096 | 0.03 GiB | — | — |
| 8,192 | 0.06 GiB | — | — |
| 16,384 | 0.12 GiB | — | — |
| 32,768 | 0.23 GiB | — | — |
| 65,536 | 0.47 GiB | — | — |
| 131,072 | 0.94 GiB | — | — |