Function createReranker

createReranker(
modelPath: string,
opts?: { nCtx?: number; nSeqMax?: number },
): Promise<Reranker>
Create a Reranker backed by a dedicated reranking model context

Loads a separate model (typically a cross-encoder) into its own KV cache and exposes score, tokenizeChunks, and dispose methods. The returned score method yields ScoredResult batches as an async iterable, mapping raw indices back to the original Chunk metadata.
Parameters
- modelPath: string
  Absolute path to the reranking model file (GGUF)
- Optionalopts: { nCtx?: number; nSeqMax?: number }
  Optional context sizing overrides
  - OptionalnCtx?: number
    Context window size for the reranker model (default 4096)
  - OptionalnSeqMax?: number
    Maximum parallel scoring sequences (default 8)
Returns Promise<Reranker>
A ready-to-use reranker instance; call dispose() when finished
- Defined in rig/src/reranker.ts:23