Is Gemma-The-Writer-N-Restless-Quill-10B-Uncensored 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 1.6x smaller, because of how its attention layers cache.

Model comparison

Gemma-The-Writer-N-Restless-Quill-10B-Uncensored 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 1.6× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Gemma-The-Writer-N-Restless-Quill-10B-Uncensored	llama-3-youko-8b
Parameters	10.0B	8.0B
Architecture	gemma2	llama
Layers	46	32
Native context	8,192	8,192
Mixture of experts	no	no
Quantizations published	31	2
Smallest quantization	2.37 GiB	5.34 GiB
Q4_K_M	5.68 GiB	—
Licence	apache-2.0	llama3

KV cache by context

the term that decides long-context viability

Context	Gemma-The-Writer-N-Restless-Quill-10B-Uncensored	llama-3-youko-8b	Ratio
4,096	1.44 GiB	0.50 GiB	2.88×
8,192	2.25 GiB	1.00 GiB	2.25×
16,384	3.68 GiB	2.00 GiB	1.84×
32,768	6.56 GiB	4.00 GiB	1.64×
65,536	12.31 GiB	8.00 GiB	1.54×
131,072	23.81 GiB	16.00 GiB	1.49×