Model comparison
Hunyuan-1.8B-Instruct vs HyperCLOVAX-SEED-Text-Instruct-1.5B
At Q4_K_M, Hunyuan-1.8B-Instruct is the smaller download — 1,133,084,864 bytes against 1,133,974,368.
From the file· summed bytes, KV per layer
Side by side
| Hunyuan-1.8B-Instruct | HyperCLOVAX-SEED-Text-Instruct-1.5B | |
|---|---|---|
| Parameters | 1.8B | 1.6B |
| Architecture | hunyuan-dense | llama |
| Layers | 32 | — |
| Native context | 262,144 | — |
| Mixture of experts | no | no |
| Quantizations published | 23 | 1 |
| Smallest quantization | 0.65 GiB | 1.06 GiB |
| Q4_K_M | 1.06 GiB | 1.06 GiB |
| Licence | — | other |
KV cache by context
the term that decides long-context viability
| Context | Hunyuan-1.8B-Instruct | HyperCLOVAX-SEED-Text-Instruct-1.5B | Ratio |
|---|---|---|---|
| 4,096 | 0.25 GiB | — | — |
| 8,192 | 0.50 GiB | — | — |
| 16,384 | 1.00 GiB | — | — |
| 32,768 | 2.00 GiB | — | — |
| 65,536 | 4.00 GiB | — | — |
| 131,072 | 8.00 GiB | — | — |