trunk/3afa7072db9835cb0996cb5e16279b4a51c3750c: [Inductor] Bound incremental autotuning dispatches (#196501) (#196501)
- PyTorch 近 90 天出现 2083 次
- PyTorch 近 90 天第 2054 次发版
- 上一次：同一天稍早 · trunk/81f41c172b3ca4628d9de18f7df2a518a65b90c5: C++ helpers for SymInt-aware elementwise binary meta kernels (#198875)
发生了什么
Summary: Add a state-wide dispatch budget for incremental autotuning. incremental_autotune_max_dispatches defaults to 200 . Override it through torch.compile() options or TORCHINDUCTOR_INCREMENTAL_AUTOTUNE_MAX_DISPATCHES ; 0 disables the cap, and negative values are rejected. When an enabled budget is exhausted, freeze the current best launcher immediately. Pending timing events still drain but cannot change the sel…
摘要按规则整理自下方来源原文