Holo-3.1-0.8B

18 of 24 layers use linear attention, which keeps a fixed-size recurrent state instead of a per-token cache. Those layers do not grow with context at all — treating them as ordinary attention, as a flat formula does, overstates this model's cache by roughly 4.0× at long context.

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 0.58 GiB. The real file is 0.63 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

How much VRAM does Holo-3.1-0.8B need?

Q4_K_M is exactly 672,330,784 bytes (0.63 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.

How large is Holo-3.1-0.8B's KV cache?

0.38 GiB at 32K context with an f16 cache, computed per layer. Quantizing the cache to q8_0 roughly halves it, which is often the difference between a context length fitting and not.

Which quantization of Holo-3.1-0.8B 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.