Open Models Almanac
Methodsinference & servingdecoding strategy

implementation detail · filed under inference & serving

Standardized sampling configuration

Applies one shared sampling configuration across use cases; some evidence specifies temperature, top-p, and top-k values.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See decoding strategy for the methods filed beside it.

sources
3
model
1
lab adopt it
1
strongest
default

Documented in

Google DeepMind

Evidence

3 quotes from 3 sources, one card per model, each opening with its strongest.

How sources treat it
used 1default 2

Gemma 4 31B

Google DeepMindadopted3 sources

Use the following standardized sampling configuration across all use cases: - temperature=1.0 - top_p=0.95 - top_k=64

defaultinference servingin Gemma 4Google DeepMind
2 more from this model

Use the following standardized sampling configuration across all use cases: - temperature=1.0 - top_p=0.95 - top_k=64

defaultinference servingin Gemma 4Google DeepMind

Use the following standardized sampling configuration across all use cases:

usedinference servingin Gemma 4Google DeepMind
↑