Second text lineage: nvidia/nemotron-3-super-120b-a12b - #421

Draft

sebasmos wants to merge 33 commits into

mainfrom

feat/text-lane-model-flag

Commits on Sep 5, 2026

Text lane: --model flag, per-model key and backend dispatch, second-lineage blind-metric arm
sebasmos
committed
228101c
Extend the nemotron-3-super-120b-a12b blind-metric arm to n=100
sebasmos
committed
bea8662
Extend the Gemini comparator to n=100 so both lineages run the same 100 cases
sebasmos
committed
448b372
Share the text-lane model dispatch and port all nineteen single-model text runners to it
sebasmos
committed
adf61fc
Pace second-vendor calls and survive a rate-limit stall
sebasmos
committed
a493599
Run the four paper-headline MedQA arms on nemotron-3-super-120b-a12b at n=120
sebasmos
committed
19ca815

Commits on Sep 6, 2026

Commits on Sep 7, 2026

Commits on Sep 8, 2026

Commits on Sep 9, 2026

Commits on Sep 10, 2026

Commits on Sep 12, 2026

Commits on Sep 13, 2026

Ignore the rebuilt MIMIC-CXR cohort manifests by their actual filenames
sebasmos
committed
7a75f7c