10 Vocabs/ˈɪnfərəns/
What does inference mean?
the process of running a trained model to get predictions or outputs
Examples in context
- Inference is fast on the new GPU.
- We run inference in real time.
- Inference latency dropped by 40%.
- We serve 10K inferences per second.
A distinction worth remembering
In everyday English, an inference is a conclusion drawn from evidence. In AI, inference means using an already trained model to produce an output.
Usage tip
Use when the model is generating outputs.
Try it yourself
Explain the difference between training a model and using it to answer a question.
Explore this vocabulary topic · More words · Read current briefings