pub fn run_benchmark(
corpus: &FibBenchmarkCorpus,
quantizer: &FibQuantizer,
) -> Result<FibBenchmarkReceiptV1>Expand description
Run a full benchmark on a corpus with a FibQuantizer.
Encodes all corpus vectors, decodes them, measures reconstruction quality, then evaluates recall@K and nDCG@K for each query against the ground truth. Timing covers the encode and decode phases separately.