🐛 Describe the bug

PyTorch-2.1.2 is a big release with a lots of new feature, so we need to make sure that:

• CUDA
  • pypi binaries with slimmed dependencies are usable in standard AWS containers (amazonlinux:2 regression in 1.13)
  • pypi binaries with slimmed dependencies are usable with stock Ubuntu-20.04: torch-2.0.0-rc1 and torch-1.13.1 can not be installed on Ubuntu 20.04  #91067 . Test: https://github.com/pytorch/builder/actions/runs/7174696263/job/19536662484#step:11:3259
  • PyTorch+CUDA11.8 and CUDA 12.1 has a working FFT on 4090
  • Check cuda 1.12.1 update issue: torch.linalg.eigh fails on GPU #94772 with small wheels
• torch.compile
  • Basic test works (for example see test mentioned in Search for libdevice relative to shared library triton-lang/triton#1176 ) in PyTorch docker container
  • torch.compile produces a binary which can be used on 3090
  • torch.compile raises an error if used on Windows. Test: https://github.com/pytorch/builder/actions/runs/7118825750/job/19382609324#step:9:12935
  • torch.compile works on 3.11 : Test: https://github.com/pytorch/builder/actions/runs/7118825750/job/19382623218#step:11:16412
• MPS
  • Resnet is usable out of the box (i.e. https://github.com/pytorch/builder/blob/main/test/smoke_test/smoke_test.py passes for MPS device). Test: https://github.com/pytorch/builder/actions/runs/7118825750/job/19382643599#step:9:301
• Validate docker release builds

Versions

2.1.2