Model comparison
Qwen2-0.5B 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
| Qwen2-0.5B | ced-base | |
|---|---|---|
| Parameters | 494M | 86M |
| Architecture | qwen2 | ced |
| Layers | 24 | — |
| Native context | 131,072 | — |
| Mixture of experts | no | no |
| Quantizations published | 31 | 3 |
| Smallest quantization | 0.29 GiB | 0.12 GiB |
| Q4_K_M | 0.37 GiB | — |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | Qwen2-0.5B | ced-base | Ratio |
|---|---|---|---|
| 4,096 | 0.05 GiB | — | — |
| 8,192 | 0.09 GiB | — | — |
| 16,384 | 0.19 GiB | — | — |
| 32,768 | 0.38 GiB | — | — |
| 65,536 | 0.75 GiB | — | — |
| 131,072 | 1.50 GiB | — | — |