Is Chocolatine-3B-Instruct-DPO-Revised or gemma-4-E4B-it smaller?

Chocolatine-3B-Instruct-DPO-Revised is smaller at Q4_K_M: 2.23 GiB against 4.64 GiB.

Which uses less memory at long context?

gemma-4-E4B-it — its KV cache at 32K context is 23.7x smaller, because of how its attention layers cache.

Model comparison

Chocolatine-3B-Instruct-DPO-Revised vs gemma-4-E4B-it

At Q4_K_M, Chocolatine-3B-Instruct-DPO-Revised is the smaller download — 2,393,231,840 bytes against 4,977,171,584. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 23.7× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Chocolatine-3B-Instruct-DPO-Revised	gemma-4-E4B-it
Parameters	3.8B	8.0B
Architecture	phi3	gemma4
Layers	32	42
Native context	4,096	131,072
Mixture of experts	no	no
Quantizations published	1	36
Smallest quantization	2.23 GiB	3.30 GiB
Q4_K_M	2.23 GiB	4.64 GiB
Licence	mit	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Chocolatine-3B-Instruct-DPO-Revised	gemma-4-E4B-it	Ratio
4,096	1.50 GiB	0.12 GiB	12.19×
8,192	3.00 GiB	0.18 GiB	16.88×
16,384	6.00 GiB	0.29 GiB	20.90×
32,768	12.00 GiB	0.51 GiB	23.72×
65,536	24.00 GiB	0.94 GiB	25.44×
131,072	48.00 GiB	1.82 GiB	26.40×