Model comparison
mxbai-embed-xsmall-v1 vs all-MiniLM-L6-v2
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| mxbai-embed-xsmall-v1 | all-MiniLM-L6-v2 | |
|---|---|---|
| Parameters | 24M | 23M |
| Architecture | bert | bert |
| Layers | 6 | 6 |
| Native context | 4,096 | 512 |
| Mixture of experts | no | no |
| Quantizations published | 4 | 28 |
| Smallest quantization | 0.03 GiB | 0.02 GiB |
| Q4_K_M | — | 0.02 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | mxbai-embed-xsmall-v1 | all-MiniLM-L6-v2 | Ratio |
|---|---|---|---|
| 4,096 | 0.04 GiB | 0.04 GiB | 1.00× |
| 8,192 | 0.07 GiB | 0.07 GiB | 1.00× |
| 16,384 | 0.14 GiB | 0.14 GiB | 1.00× |
| 32,768 | 0.28 GiB | 0.28 GiB | 1.00× |
| 65,536 | 0.56 GiB | 0.56 GiB | 1.00× |
| 131,072 | 1.13 GiB | 1.13 GiB | 1.00× |