Model comparison

polyglot-lion-1.7b vs Qwen3-4B

At Q4_K_M, polyglot-lion-1.7b is the smaller download — 1,107,408,160 bytes against 2,497,280,256.

From the file· summed bytes, KV per layer

Side by side

polyglot-lion-1.7bQwen3-4B
Parameters2.0B4.0B
Architectureqwen3vlqwen3
Layers36
Native context40,960
Mixture of expertsnono
Quantizations published3629
Smallest quantization0.48 GiB1.01 GiB
Q4_K_M1.03 GiB2.33 GiB
Licencemit

KV cache by context

the term that decides long-context viability
Contextpolyglot-lion-1.7bQwen3-4BRatio
4,0960.56 GiB
8,1921.13 GiB
16,3842.25 GiB
32,7684.50 GiB
65,5369.00 GiB
131,07218.00 GiB