Model comparison
embeddinggemma-300m-qat-q4_0-unquantized 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
| embeddinggemma-300m-qat-q4_0-unquantized | ced-base | |
|---|---|---|
| Parameters | 303M | 86M |
| Architecture | gemma-embedding | ced |
| Layers | 24 | — |
| Native context | 2,048 | — |
| Mixture of experts | no | no |
| Quantizations published | 2 | 3 |
| Smallest quantization | 0.21 GiB | 0.12 GiB |
| Q4_K_M | — | — |
| Licence | gemma | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | embeddinggemma-300m-qat-q4_0-unquantized | ced-base | Ratio |
|---|---|---|---|
| 4,096 | 0.04 GiB | — | — |
| 8,192 | 0.05 GiB | — | — |
| 16,384 | 0.08 GiB | — | — |
| 32,768 | 0.14 GiB | — | — |
| 65,536 | 0.27 GiB | — | — |
| 131,072 | 0.52 GiB | — | — |