Throughput

Prediction

भविष्यवाणी / अनुमान

The output a model produces for a given input, such as a label, number, or generated text.

Rate Limit

रेट लिमिट / दर सीमा

A cap on how many requests a user can make to a service in a time window.

AUC

एयूसी

The area under the ROC curve summarizing classification performance as a single number.

Autoregressive

अटोरिग्रेसिभ

Generating output one token at a time, each based on all the tokens produced before it.

Batch Inference

ब्याच इन्फरेन्स

Running predictions on many inputs together rather than one at a time.

Cold Start

कोल्ड स्टार्ट

The delay or difficulty when a system or model runs for the first time after being idle.

Collaborative Filtering

सहयोगी छनोट

Recommending items by finding users with similar tastes and patterns.

Confidence Score

विश्वास अंक

A number expressing how certain a model is about a particular prediction.