nx_qabench_corpus_fetch.nx
buildroot/runtime/nx_qabench_corpus_fetch.nx
about
nx_qabench_corpus_fetch.nx -- GROW the PPMI corpus ~6x with rows DISJOINT from the eval subsets (eval =
squad/hotpot pages 0-2, musique pages 0-1; corpus chunks start BEYOND those). PPMI quality scales with
corpus; the ladder eval sets stay untouched = a clean corpus-scaling datapoint vs the 161/153/37 baseline.
Chunks of 5 pages (500 rows, ~3-6MB each, under the 16MB reader cap). Idempotent + atomic per chunk.
Run SOLO. expect_exit: 0 license_tier: ORIGINAL
dependencies 1 imports · 0 importers
imports: nx_research_fetch_ext.nx
imported by: nobody (leaf or entry point)
call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown
structs
| none |
consts
| none |
functions
| 8 | func main() -> i64 calls 1: trust_store_count |