Model comparison
Lightning-1.7B vs HyperCLOVAX-SEED-Text-Instruct-1.5B
At Q4_K_M, Lightning-1.7B is the smaller download — 1,107,408,960 bytes against 1,133,974,368.
From the file· summed bytes, KV per layer
Side by side
| Lightning-1.7B | HyperCLOVAX-SEED-Text-Instruct-1.5B | |
|---|---|---|
| Parameters | 1.7B | 1.6B |
| Architecture | qwen3 | llama |
| Layers | 28 | — |
| Native context | 40,960 | — |
| Mixture of experts | no | no |
| Quantizations published | 23 | 1 |
| Smallest quantization | 0.65 GiB | 1.06 GiB |
| Q4_K_M | 1.03 GiB | 1.06 GiB |
| Licence | mpl-2.0 | other |
KV cache by context
the term that decides long-context viability
| Context | Lightning-1.7B | HyperCLOVAX-SEED-Text-Instruct-1.5B | 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 | — | — |