Model comparison

Fourier-Qwen2.5-VL-3B-0.67 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

Fourier-Qwen2.5-VL-3B-0.67HyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters3.8B1.6B
Architectureqwen2vlllama
Layers36
Native context128,000
Mixture of expertsnono
Quantizations published241
Smallest quantization0.74 GiB1.06 GiB
Q4_K_M1.06 GiB
Licenceapache-2.0other

KV cache by context

the term that decides long-context viability
ContextFourier-Qwen2.5-VL-3B-0.67HyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,0960.14 GiB
8,1920.28 GiB
16,3840.56 GiB
32,7681.13 GiB
65,5362.25 GiB
131,0724.50 GiB