Is deepseek-math-7b-instruct or llama-3-youko-8b smaller?

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

Which uses less memory at long context?

llama-3-youko-8b — its KV cache at 32K context is 3.8x smaller, because of how its attention layers cache.

Model comparison

deepseek-math-7b-instruct vs llama-3-youko-8b

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: llama-3-youko-8b's KV cache at 32K is 3.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	deepseek-math-7b-instruct	llama-3-youko-8b
Parameters	6.9B	8.0B
Architecture	llama	llama
Layers	30	32
Native context	4,096	8,192
Mixture of experts	no	no
Quantizations published	38	2
Smallest quantization	2.53 GiB	5.34 GiB
Q4_K_M	3.93 GiB	—
Licence	other	llama3

KV cache by context

the term that decides long-context viability

Context	deepseek-math-7b-instruct	llama-3-youko-8b	Ratio
4,096	1.88 GiB	0.50 GiB	3.75×
8,192	3.75 GiB	1.00 GiB	3.75×
16,384	7.50 GiB	2.00 GiB	3.75×
32,768	15.00 GiB	4.00 GiB	3.75×
65,536	30.00 GiB	8.00 GiB	3.75×
131,072	60.00 GiB	16.00 GiB	3.75×