Model comparison
dolphin-2_6-phi-2 vs HyperCLOVAX-SEED-Text-Instruct-1.5B
At Q4_K_M, HyperCLOVAX-SEED-Text-Instruct-1.5B is the smaller download — 1,133,974,368 bytes against 1,789,240,608.
From the file· summed bytes, KV per layer
Side by side
| dolphin-2_6-phi-2 | HyperCLOVAX-SEED-Text-Instruct-1.5B | |
|---|---|---|
| Parameters | 2.8B | 1.6B |
| Architecture | phi2 | llama |
| Layers | 32 | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 12 | 1 |
| Smallest quantization | 1.09 GiB | 1.06 GiB |
| Q4_K_M | 1.67 GiB | 1.06 GiB |
| Licence | other | other |
KV cache by context
the term that decides long-context viability
| Context | dolphin-2_6-phi-2 | HyperCLOVAX-SEED-Text-Instruct-1.5B | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |