@promptbook/azure-openai
Version:
Promptbook: Run AI apps in plain human language across multiple models and platforms
57 lines (56 loc) • 1.92 kB
TypeScript
import { ExpectError } from '../../errors/ExpectError';
import type { Prompt } from '../../types/Prompt';
import type { ChatPromptResult } from '../PromptResult';
import type { CompletionPromptResult } from '../PromptResult';
import type { EmbeddingPromptResult } from '../PromptResult';
import type { PromptResult } from '../PromptResult';
/**
* Represents the ongoing result of a pipeline task execution
*
* Used internally by the pipeline executor to track state during execution attempts
*
* @private internal utility of `createPipelineExecutor`
*/
export type $OngoingTaskResult = {
/**
* The prompt object used for the current execution attempt, if applicable.
*/
$prompt?: Prompt;
/**
* The result of a chat model execution, if applicable.
*/
$chatResult?: ChatPromptResult;
/**
* The result of a completion model execution, if applicable.
*/
$completionResult?: CompletionPromptResult;
/**
* The result of an embedding model execution, if applicable.
* Note: [🤖] Embedding results are less common in standard pipelines.
*/
$embeddingResult?: EmbeddingPromptResult;
/**
* The final result object for the task, or null if not yet available.
*/
$result: PromptResult | null;
/**
* The result string produced by the task, or null if not yet available.
*/
$resultString: string | null;
/**
* The last expectation error encountered, or null if none.
*/
$expectError: ExpectError | null;
/**
* List of errors encountered during script postprocessing or execution.
*/
$scriptPipelineExecutionErrors: Array<Error>;
/**
* Array of all failed attempts, storing both the result string and the error for each failure
*/
$failedResults: Array<{
attemptIndex: number;
result: string | null;
error: ExpectError;
}>;
};