Model comparison
SmolVLM2-256M-Video-Instruct vs SmolLM2-135M-Instruct
These two publish different quantization sets; the table below has the exact sizes.
From the file· summed bytes, KV per layer
Side by side
| SmolVLM2-256M-Video-Instruct | SmolLM2-135M-Instruct | |
|---|---|---|
| Parameters | 256M | 135M |
| Architecture | llama | llama |
| Layers | 30 | 30 |
| Native context | 8,192 | 8,192 |
| Mixture of experts | no | no |
| Quantizations published | 26 | 41 |
| Smallest quantization | 0.09 GiB | 0.08 GiB |
| Q4_K_M | — | 0.10 GiB |
| Licence | apache-2.0 | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | SmolVLM2-256M-Video-Instruct | SmolLM2-135M-Instruct | Ratio |
|---|---|---|---|
| 4,096 | 0.09 GiB | 0.09 GiB | 1.00× |
| 8,192 | 0.18 GiB | 0.18 GiB | 1.00× |
| 16,384 | 0.35 GiB | 0.35 GiB | 1.00× |
| 32,768 | 0.70 GiB | 0.70 GiB | 1.00× |
| 65,536 | 1.41 GiB | 1.41 GiB | 1.00× |
| 131,072 | 2.81 GiB | 2.81 GiB | 1.00× |