Model comparison
Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliterated 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 3,416,119,296.
From the file· summed bytes, KV per layer
Side by side
| Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliterated | HyperCLOVAX-SEED-Text-Instruct-1.5B | |
|---|---|---|
| Parameters | 5.1B | 1.6B |
| Architecture | gemma4 | llama |
| Layers | 35 | — |
| Native context | 131,072 | — |
| Mixture of experts | no | no |
| Quantizations published | 28 | 1 |
| Smallest quantization | 2.16 GiB | 1.06 GiB |
| Q4_K_M | 3.18 GiB | 1.06 GiB |
| Licence | apache-2.0 | other |
KV cache by context
the term that decides long-context viability
| Context | Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliterated | HyperCLOVAX-SEED-Text-Instruct-1.5B | Ratio |
|---|---|---|---|
| 4,096 | 0.05 GiB | — | — |
| 8,192 | 0.08 GiB | — | — |
| 16,384 | 0.14 GiB | — | — |
| 32,768 | 0.25 GiB | — | — |
| 65,536 | 0.46 GiB | — | — |
| 131,072 | 0.90 GiB | — | — |