Is Qwen2.5-Coder-1.5B-Instruct-abliterated or Llama-3.2-1B-Instruct smaller?

Llama-3.2-1B-Instruct is smaller at Q4_K_M: 0.75 GiB against 1.04 GiB.

Which uses less memory at long context?

Qwen2.5-Coder-1.5B-Instruct-abliterated — its KV cache at 32K context is 1.1x smaller, because of how its attention layers cache.

Model comparison

At Q4_K_M, Llama-3.2-1B-Instruct is the smaller download — 807,690,688 bytes against 1,117,321,696.

From the file· summed bytes, KV per layer

	Qwen2.5-Coder-1.5B-Instruct-abliterated	Llama-3.2-1B-Instruct
Parameters	1.8B	1.2B
Architecture	qwen2	llama
Layers	28	16
Native context	32,768	131,072
Mixture of experts	no	no
Quantizations published	35	39
Smallest quantization	0.65 GiB	0.39 GiB
Q4_K_M	1.04 GiB	0.75 GiB
Licence	apache-2.0	—

the term that decides long-context viability

Context	Qwen2.5-Coder-1.5B-Instruct-abliterated	Llama-3.2-1B-Instruct	Ratio
4,096	0.11 GiB	0.13 GiB	1.14×
8,192	0.22 GiB	0.25 GiB	1.14×
16,384	0.44 GiB	0.50 GiB	1.14×
32,768	0.88 GiB	1.00 GiB	1.14×
65,536	1.75 GiB	2.00 GiB	1.14×
131,072	3.50 GiB	4.00 GiB	1.14×