Track: research · Level: core · Effort: ~10h (6h GPU) · Depends on: #335 (T20)
Why this matters
Flowers102 has only ten images per class, an extreme low-data regime. If BNNR helps here, it strengthens the case that saliency-guided augmentation matters most exactly when data is scarce, which is the situation most practitioners actually face.
Steps
- Get maintainer sign-off on the run plan.
- Run python benchmarks/run_grand_benchmark.py --dataset flowers102 --device cuda.
- Back up the JSON and overlays; sanity-check the numbers.
Done when
Complete flowers102 JSON; numbers sane; handed to T35.
How to take this: comment "taking this" and wait to be assigned. Branch t31-short-desc from upstream/main, and put Closes #<this issue number> in your PR. Full workflow: the Cohort Handbook (pinned in Discord).
Track: research · Level: core · Effort: ~10h (6h GPU) · Depends on: #335 (T20)
Why this matters
Flowers102 has only ten images per class, an extreme low-data regime. If BNNR helps here, it strengthens the case that saliency-guided augmentation matters most exactly when data is scarce, which is the situation most practitioners actually face.
Steps
Done when
Complete flowers102 JSON; numbers sane; handed to T35.
How to take this: comment "taking this" and wait to be assigned. Branch
t31-short-descfromupstream/main, and putCloses #<this issue number>in your PR. Full workflow: the Cohort Handbook (pinned in Discord).