Model comparison
HunyuanOCR 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
| HunyuanOCR | ced-base | |
|---|---|---|
| Parameters | 1.1B | 86M |
| Architecture | hunyuan_vl | ced |
| Layers | 24 | — |
| Native context | 131,072 | — |
| Mixture of experts | yes, 1 experts | no |
| Quantizations published | 25 | 3 |
| Smallest quantization | 0.24 GiB | 0.12 GiB |
| Q4_K_M | 0.33 GiB | — |
| Licence | — | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | HunyuanOCR | ced-base | Ratio |
|---|---|---|---|
| 4,096 | 0.11 GiB | — | — |
| 8,192 | 0.21 GiB | — | — |
| 16,384 | 0.42 GiB | — | — |
| 32,768 | 0.84 GiB | — | — |
| 65,536 | 1.69 GiB | — | — |
| 131,072 | 3.38 GiB | — | — |