@ai-foundry/llm-sdk/dist/models/llm.models.d.ts

This is just another SDK for the common LLM API providers.

import type { JSONObject } from './data.models';
export type FinishReason = 'stop' | 'length' | 'content-filter' | 'tool-calls' | 'error' | 'other' | 'unknown';
export interface LLMTokensUsage {
    promptTokens: number;
    completionTokens: number;
    totalTokens: number;
}
export interface LLMRequest {
    body: JSONObject;
    headers: Record<string, string>;
}
export interface LLMResponse {
    body: JSONObject;
    headers: Record<string, string>;
}
export interface LLMApiProviderOptions {
    headers?: Record<string, string>;
}
export interface LLMOptions {
    /**
     * Maximum number of tokens to generate.
     */
    maxTokens?: number;
    /**
     * Temperature setting.
     * The value is passed through to the provider. The range depends on the provider and model.
     * It is recommended to set either `temperature` or `topP`, but not both.
     */
    temperature?: number;
    /**
     * Nucleus sampling.
     * The value is passed through to the provider. The range depends on the provider and model.
     * It is recommended to set either `temperature` or `topP`, but not both.
     */
    topP?: number;
    /**
     * Only sample from the top K options for each subsequent token.
     * Used to remove "long tail" low probability responses.
     * Recommended for advanced use cases only. You usually only need to use temperature.
     */
    topK?: number;
    /**
     * Frequency penalty setting.
     * It affects the likelihood of the model to repeatedly use the same words or phrases.
     * The value is passed through to the provider. The range depends on the provider and model.
     */
    frequencyPenalty?: number;
    /**
     * Presence penalty setting.
     * It affects the likelihood of the model to repeat information that is already in the prompt.
     * The value is passed through to the provider. The range depends on the provider and model.
     */
    presencePenalty?: number;
    /**
     * Stop sequences.
     * If set, the model will stop generating text when one of the stop sequences is generated.
     */
    stopSequences?: string[];
    /**
     * The seed (integer) to use for random sampling.
     * If set and supported by the model, calls will generate deterministic results.
     */
    seed?: number;
}