foxhunt

Files

jgrusewski 25ec8c7bcf perf(cuda): fuse 5 label gathers into sample_and_gather — 95.6% GPU time eliminated

nsys profile showed gpu_gather_bce_labels consumed 95.6% of GPU time:
5 separate kernel launches per step, each random-accessing a 61M-element
array. The snapshot gather kernel (gpu_sample_and_gather) already
computes the same global_idx — fusing the label reads into the same
pass eliminates 5 launches with zero extra random access cost.

Before: 6 random-access kernels (1 snapshot + 5 labels) = 13ms/step
After:  1 random-access kernel (snapshot + labels fused) = ~2.5ms/step

Expected speedup: 6 sps → 30-50+ sps at b=1024.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-05-27 21:14:26 +02:00

cuda

perf(cuda): fuse 5 label gathers into sample_and_gather — 95.6% GPU time eliminated

2026-05-27 21:14:26 +02:00

examples

feat(rl): wire checkpoint save/resume into training loop

2026-05-27 20:30:50 +02:00

src

perf(cuda): fuse 5 label gathers into sample_and_gather — 95.6% GPU time eliminated

2026-05-27 21:14:26 +02:00

tests

perf(rl): fuse adamw_increment into adamw_step + tau into iqn_forward