Model comparison
SmolLM-1.7B-Instruct-v0.2 vs Llama-3.2-3B-Instruct
At Q4_K_M, SmolLM-1.7B-Instruct-v0.2 is the smaller download — 1,055,610,144 bytes against 2,019,377,440.
From the file· summed bytes, KV per layer
Side by side
| SmolLM-1.7B-Instruct-v0.2 | Llama-3.2-3B-Instruct | |
|---|---|---|
| Parameters | 1.7B | 3.2B |
| Architecture | llama | llama |
| Layers | — | 28 |
| Native context | — | 131,072 |
| Mixture of experts | no | no |
| Quantizations published | 32 | 39 |
| Smallest quantization | 0.38 GiB | 0.85 GiB |
| Q4_K_M | 0.98 GiB | 1.88 GiB |
| Licence | apache-2.0 | — |
KV cache by context
the term that decides long-context viability
| Context | SmolLM-1.7B-Instruct-v0.2 | Llama-3.2-3B-Instruct | Ratio |
|---|---|---|---|
| 4,096 | — | 0.44 GiB | — |
| 8,192 | — | 0.88 GiB | — |
| 16,384 | — | 1.75 GiB | — |
| 32,768 | — | 3.50 GiB | — |
| 65,536 | — | 7.00 GiB | — |
| 131,072 | — | 14.00 GiB | — |