Function createOpenAiEmbedder

packages/server/src/embeddings/embedder.ts:128–274

(
  config: OpenAiEmbedderConfig,
  options: OpenAiEmbedderOptions = {},
)

}

export function createOpenAiEmbedder(
  config: OpenAiEmbedderConfig,
  options: OpenAiEmbedderOptions = {},
): Embedder {
  const fetchImpl = options.fetchImpl ?? fetch;
  const sleep = options.sleep ?? defaultSleep;
  const maxBatchSize = options.maxBatchSize ?? DEFAULTS.maxBatchSize;
  const maxBatchChars = options.maxBatchChars ?? DEFAULTS.maxBatchChars;
  const maxRetries = options.maxRetries ?? DEFAULTS.maxRetries;
  const backoffBaseMs = options.backoffBaseMs ?? DEFAULTS.backoffBaseMs;
  const docTimeoutMs = options.docTimeoutMs ?? DEFAULTS.docTimeoutMs;
  const queryTimeoutMs = options.queryTimeoutMs ?? DEFAULTS.queryTimeoutMs;

  assertSafeEmbeddingsBaseUrl(config.baseUrl);
  const dims = config.dimensions ?? DEFAULT_EMBEDDINGS_DIMENSIONS;
  const endpoint = `${config.baseUrl.replace(/\/+$/, '')}/embeddings`;

  function batchInputs(texts: readonly string[]): string[][] {
    const batches: string[][] = [];
    let current: string[] = [];
    let chars = 0;
    for (const t of texts) {
      if (
        current.length > 0 &&
        (current.length >= maxBatchSize || chars + t.length > maxBatchChars)
      ) {
        batches.push(current);
        current = [];
        chars = 0;
      }
      current.push(t);
      chars += t.length;
    }
    if (current.length > 0) batches.push(current);
    return batches;
  }

  type AttemptResult =
    | { kind: 'ok'; vectors: Float32Array[] }
    | { kind: 'retry'; reason: EmbeddingErrorReason; error: Error }
    | { kind: 'fatal'; reason: EmbeddingErrorReason; error: Error };

  async function attemptOnce(
    body: string,
    expectedCount: number,
    roleLabel: 'query' | 'document',
    timeoutMs: number,
  ): Promise<AttemptResult> {
    const controller = new AbortController();
    const timer = setTimeout(() => controller.abort(), timeoutMs);
    const startedAt = performance.now();
    try {
      const res = await fetchImpl(endpoint, {
        method: 'POST',
        headers: {
          'Content-Type': 'application/json',
          Authorization: `Bearer ${config.apiKey}`,
        },