InferenceDetails interface
Details for an inference call.
Properties
| endpoint | Endpoint the model is hosted on. |
| finish |
Finish reasons returned by the model. |
| input |
Number of input (prompt) tokens. |
| model | Name of the model used for inference. |
| operation |
The inference operation type (e.g. chat, text completion). |
| output |
Number of output (completion) tokens. |
| provider |
Name of the provider serving the model. |
| thought |
The model's thought process / reasoning trace. |
Property Details
endpoint
finishReasons
Finish reasons returned by the model.
finishReasons?: string[]
Property Value
string[]
inputTokens
Number of input (prompt) tokens.
inputTokens?: number
Property Value
number
model
Name of the model used for inference.
model: string
Property Value
string
operationName
The inference operation type (e.g. chat, text completion).
operationName: InferenceOperationType
Property Value
outputTokens
Number of output (completion) tokens.
outputTokens?: number
Property Value
number
providerName
Name of the provider serving the model.
providerName?: string
Property Value
string
thoughtProcess
The model's thought process / reasoning trace.
thoughtProcess?: string
Property Value
string