This jobs is failing on trunk with the following error https://github.com/pytorch/pytorch/actions/runs/12025459038/job/33524520375#step:21:628.

After chatting with @yanboliang and @eellison, the failure is related to cudagraphs but there is not an easy fix on the cudagraphs side and would need to be fixed upstream in the model. The failing model is Llama-2-7b-chat-hf from huggingface.

cc @seemethere @malfet @pytorch/pytorch-dev-infra @chauhang @penguinwu @voznesenskym @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @aakhundov