Model comparison

gemma-2-2b-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 1,708,582,784.

From the file· summed bytes, KV per layer

Side by side

gemma-2-2b-it-abliteratedHyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters2.6B1.6B
Architecturegemma2llama
Layers26
Native context8,192
Mixture of expertsnono
Quantizations published291
Smallest quantization1.15 GiB1.06 GiB
Q4_K_M1.59 GiB1.06 GiB
Licencegemmaother

KV cache by context

the term that decides long-context viability
Contextgemma-2-2b-it-abliteratedHyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,0960.41 GiB
8,1920.63 GiB
16,3841.04 GiB
32,7681.85 GiB
65,5363.48 GiB
131,0726.73 GiB
gemma-2-2b-it-abliterated vs HyperCLOVAX-SEED-Text-Instruct-1.5B — size, memory and hardware fit — ossmodeldb