-
da22885645
Add multi-op batch screening layer; fix cudagraph fallback regressions
main
zhoulin
2026-08-12 19:04:09 +00:00
-
95895cea0e
Adapt to upstream FlagGems repo split and record pre_hook configs for libtuner
zhoulin
2026-08-12 10:55:32 +00:00
-
1e1d612031
Add opt-in multi-GPU sweep parallelism; fix GPU pinning and run dir races
zhoulin
2026-07-29 12:14:12 +00:00
-
7e0e8648f1
Fix autotune record/replay silently skipping libtuner kernels
zhoulin
2026-07-29 00:38:14 +00:00
-
f98ec10fd0
Renew scripts
zhoulin
2026-07-24 10:17:33 +00:00
-
c8a42b66ee
Move trace shape sets to top-level shape/ and frame ab_fold_test.sh as A/B reference example
zhoulin
2026-07-20 16:53:14 +00:00
-
00902e6c67
Default USE_FLAGTUNE to 0 and add fold_scale A/B script with trace shapes
zhoulin
2026-07-20 16:50:59 +00:00
-
26b071c6e1
Stabilize first-run cudagraph timing and colorize terminal output
zhoulin
2026-07-19 19:21:13 +00:00
-
8ace0d790c
Add FlagGems single-op perf benchmark harness
zhoulin
2026-07-16 03:52:07 +00:00
-
242d74af0e
Initial commit
zhoulin
2026-07-15 17:20:27 +08:00