Files
pytorch/torch
Shunting Zhang 40b25578e4 [Inductor] deterministic mode (#163589)
Add a deterministic mode to skip the on device benchmarking that we know should affect numeric. This include
- pad-mm
- dynamic rblock scaling
- template autotuning
- coordinate descent tuning for reduction
- reduction config autotuning in CachingAutotuner.  For reduction both RBLOCK, num_warps should affect numeric. XBLOCK does not. We can still autotune XBLOCK for reductions.
- benchmarking for computation communication reordering pass

The mode definitely has perf hit.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163589
Approved by: https://github.com/v0i0
2025-10-04 01:05:08 +00:00
..
2025-10-02 07:31:11 +00:00
2025-09-29 14:49:19 +00:00
2025-04-27 09:56:42 +00:00