Model comparison

Huihui-gemma-3n-E4B-it-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 4,237,064,192.

From the file· summed bytes, KV per layer

Side by side

Huihui-gemma-3n-E4B-it-abliteratedHyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters7.8B1.6B
Architecturegemma3nllama
Layers35
Native context32,768
Mixture of expertsnono
Quantizations published331
Smallest quantization2.57 GiB1.06 GiB
Q4_K_M3.95 GiB1.06 GiB
Licencegemmaother

KV cache by context

the term that decides long-context viability
ContextHuihui-gemma-3n-E4B-it-abliteratedHyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,0960.11 GiB
8,1920.16 GiB
16,3840.27 GiB
32,7680.49 GiB
65,5360.93 GiB
131,0721.80 GiB