mirror of
https://github.com/jung-geun/PSO.git
synced 2026-09-20 14:11:48 +09:00
Keep objective construction on CPU and evaluate cached detection features in source-sized chunks. Adam gradients now accumulate per chunk, avoiding a monolithic 2,500-image CUDA graph without changing the full-objective mean. Constraint: Preserve 2,500-image objective and exact 41 Adam evaluations Rejected: Reduce objective sample count | changes the sealed protocol Confidence: high Scope-risk: moderate Not-tested: Full remote 2,500-image CUDA optimization