Model comparison

Felldude-Uncensored-Ministral3-3B-bf16 vs HyperCLOVAX-SEED-Text-Instruct-1.5B

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

Felldude-Uncensored-Ministral3-3B-bf16HyperCLOVAX-SEED-Text-Instruct-1.5B
Parameters3.8B1.6B
Architecturemistral3llama
Layers26
Native context262,144
Mixture of expertsnono
Quantizations published241
Smallest quantization0.87 GiB1.06 GiB
Q4_K_M1.06 GiB
Licenceother

KV cache by context

the term that decides long-context viability
ContextFelldude-Uncensored-Ministral3-3B-bf16HyperCLOVAX-SEED-Text-Instruct-1.5BRatio
4,0960.41 GiB
8,1920.81 GiB
16,3841.63 GiB
32,7683.25 GiB
65,5366.50 GiB
131,07213.00 GiB
Felldude-Uncensored-Ministral3-3B-bf16 vs HyperCLOVAX-SEED-Text-Instruct-1.5B — size, memory and hardware fit — ossmodeldb