Uses torch._inductor.compile instead of torch.compile in benchmark script to avoid segmentation by Dynamo #1540
Azure Pipelines / lightning-thunder (GPUs) (testing ubuntu22.04 | cuda 12.1 | python 3.10 | torch 2.5.1 | distributed)
succeeded
Dec 11, 2024 in 18m 59s
testing ubuntu22.04 | cuda 12.1 | python 3.10 | torch 2.5.1 | distributed succeeded
Loading