Model comparison

LFM2-1.2B vs HyperCLOVAX-SEED-Text-Instruct-1.5B

At Q4_K_M, LFM2-1.2B is the smaller download — 730,893,024 bytes against 1,133,974,368.

From the file· summed bytes, KV per layer

Side by side

LFM2-1.2BHyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters1.2B1.6B
Architecturelfm2llama
Layers16
Native context128,000
Mixture of expertsnono
Quantizations published401
Smallest quantization0.45 GiB1.06 GiB
Q4_K_M0.68 GiB1.06 GiB
Licenceotherother

KV cache by context

the term that decides long-context viability
ContextLFM2-1.2BHyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,0960.13 GiB
8,1920.25 GiB
16,3840.50 GiB
32,7681.00 GiB
65,5362.00 GiB
131,0724.00 GiB