gemma-4-31B-heretic-finetune

50 of 60 layers cache only a 1,024-token window rather than the full context, on a period of . Figures assume the default configuration; --swa-full disables the saving entirely.

A parameters × bits ÷ 8 estimate puts I1-IQ1_S at roughly 16.08 GiB. The real file is 6.67 GiB, because a quantization is a mixture and some tensors are always kept at higher precision. The larger discrepancy is the cache: a flat formula gives 30.00 GiB at 32K context where the real figure is 6.17 GiB, because most of this model's layers cache a fixed window rather than the whole context.

How much VRAM does gemma-4-31B-heretic-finetune need?

I1-IQ1_S is exactly 7,156,489,696 bytes (6.67 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.

How large is gemma-4-31B-heretic-finetune's KV cache?

6.17 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 gemma-4-31B-heretic-finetune 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.