What is an Inference Result?

An inference result in Galtea represents a single turn in a conversation between a user and an AI system. It captures the user’s input and the AI’s corresponding output, along with other optional metadata such as latency and token usage.

You can create, view and manage your inference results programmatically using the Galtea SDK.

SDK Integration

The Galtea SDK allows you to create, view, and manage inference results programmatically. This is particularly useful for logging production data or for evaluating complex conversational flows.

Inference Result Properties

Session
Session
required

The session to which the inference result belongs.

Input
string
required

The input text or prompt for the inference result.

Output
string
required

The generated output or response for the inference result.

Retrieval Context
string

The context retrieved by a RAG system, if applicable.

Latency
float

The latency in milliseconds for the model’s response.

Usage Info
dict

Token usage information for the inference result.

Cost Info
dict

Cost information for the inference result.