Model comparison

Mage-Flow vs HyperCLOVAX-SEED-Text-Instruct-1.5B

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

Mage-FlowHyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters4.1B1.6B
Architecturepigllama
Layers
Native context
Mixture of expertsnono
Quantizations published11
Smallest quantization2.21 GiB1.06 GiB
Q4_K_M1.06 GiB
Licencemitother

KV cache by context

the term that decides long-context viability
ContextMage-FlowHyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,096
8,192
16,384
32,768
65,536
131,072