privacy-filter-multilingual

This model declares a 128-token sliding window, but we could not establish which layers use it. Its architecture publishes the layout as a per-layer array inside the model file rather than as a period in config.json, and we have not yet ingested that array.

A flat context × layers × heads figure would be substantially too high, so we are not showing one. This is tracked as a known gap rather than filled with a guess.

A parameters × bits ÷ 8 estimate puts F16 at roughly 0.73 GiB. The real file is 2.62 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

How much VRAM does privacy-filter-multilingual need?

F16 is exactly 2,817,460,832 bytes (2.62 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.

Is privacy-filter-multilingual a mixture-of-experts model?

Yes — 128 experts, 4 routed per token. Every expert must be resident, but only the routed ones are read per token, which is why its memory requirement and its speed behave very differently.

Which quantization of privacy-filter-multilingual should I use?

Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.