Model comparison
SmolLM2-135M-Instruct vs LFM2.5-350M
At Q4_K_M, SmolLM2-135M-Instruct is the smaller download — 105,454,144 bytes against 228,332,128. At long context the gap widens: LFM2.5-350M's KV cache at 32K is 1.9× smaller, which usually matters more than the difference in weights.
From the file· summed bytes, KV per layer
Side by side
| SmolLM2-135M-Instruct | LFM2.5-350M | |
|---|---|---|
| Parameters | 135M | 354M |
| Architecture | llama | lfm2 |
| Layers | 30 | 16 |
| Native context | 8,192 | 128,000 |
| Mixture of experts | no | no |
| Quantizations published | 41 | 13 |
| Smallest quantization | 0.08 GiB | 0.20 GiB |
| Q4_K_M | 0.10 GiB | 0.21 GiB |
| Licence | apache-2.0 | other |
KV cache by context
the term that decides long-context viability
| Context | SmolLM2-135M-Instruct | LFM2.5-350M | Ratio |
|---|---|---|---|
| 4,096 | 0.09 GiB | 0.05 GiB | 1.87× |
| 8,192 | 0.18 GiB | 0.09 GiB | 1.87× |
| 16,384 | 0.35 GiB | 0.19 GiB | 1.87× |
| 32,768 | 0.70 GiB | 0.38 GiB | 1.87× |
| 65,536 | 1.41 GiB | 0.75 GiB | 1.87× |
| 131,072 | 2.81 GiB | 1.50 GiB | 1.87× |