David Berard 9d54d3bec2 [NVFuser] undo v100 OOM skips (#90070) ... Summary: I think these were just caused by parallel tests. After adjusting test settings to 1 thread, these stopped OOMing. Test Plan: ``` $ buck2 test -j 1 mode/dev-nosan //caffe2/torch/csrc/jit/codegen/cuda:nvfuser ``` https://www.internalfb.com/intern/testinfra/testrun/6473924590389963 Differential Revision: D41643827 Pull Request resolved: https://github.com/pytorch/pytorch/pull/90070 Approved by: https://github.com/jjsjann123		2022-12-02 21:58:24 +00:00
..
cuda	[NVFuser] undo v100 OOM skips (#90070)	2022-12-02 21:58:24 +00:00
fuser	[NVFuser] Upstream push 1026 (#87779)	2022-11-04 20:04:34 +00:00
onednn	Set correct size of attr::output_layouts when the graph has multiple outputs in JIT oneDNN fuser (#88496)	2022-11-15 07:29:55 +00:00