Is csm-1b or orpheus-3b-0.1-ft smaller?

Both publish quantizations; compare the table for the exact sizes.

Which uses less memory at long context?

csm-1b — its KV cache at 32K context is 3.5x smaller, because of how its attention layers cache.

Model comparison

csm-1b vs orpheus-3b-0.1-ft

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: csm-1b's KV cache at 32K is 3.5× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	csm-1b	orpheus-3b-0.1-ft
Parameters	1.6B	3.8B
Architecture	csm	llama
Layers	16	28
Native context	2,048	131,072
Mixture of experts	no	no
Quantizations published	3	22
Smallest quantization	1.34 GiB	0.91 GiB
Q4_K_M	—	1.95 GiB
Licence	cc	apache-2.0

KV cache by context

the term that decides long-context viability

Context	csm-1b	orpheus-3b-0.1-ft	Ratio
4,096	0.13 GiB	0.44 GiB	3.50×
8,192	0.25 GiB	0.88 GiB	3.50×
16,384	0.50 GiB	1.75 GiB	3.50×
32,768	1.00 GiB	3.50 GiB	3.50×
65,536	2.00 GiB	7.00 GiB	3.50×
131,072	4.00 GiB	14.00 GiB	3.50×