[WIP] autoquant prototype

[jerryzh168]

No description provided.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1053

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 4 New Failures

As of commit 3f539d8 with merge base 3475aed ():

NEW FAILURES - The following jobs have failed:

• Run Regression Tests / test (CUDA 2.3, linux.g5.12xlarge.nvidia.gpu, torch==2.3.0, cuda, 12.1) / linux-job (gh)
  test/integration/test_integration.py::TestAutoQuant::test_autoquant_one_input_29_cuda
• Run Regression Tests / test (CUDA 2.4, linux.g5.12xlarge.nvidia.gpu, torch==2.4.0, cuda, 12.1) / linux-job (gh)
  test/integration/test_integration.py::TestAutoQuant::test_autoquant_one_input_29_cuda
• Run Regression Tests / test (CUDA 2.5, linux.g5.12xlarge.nvidia.gpu, torch==2.5.0 --index-url https://download.pytorch.o... / linux-job (gh)
  test/integration/test_integration.py::TestUtils::test_get_model_size_autoquant_5_cuda
• Run Regression Tests / test (CUDA Nightly, linux.g5.12xlarge.nvidia.gpu, --pre torch --index-url https://download.pytorc... / linux-job (gh)
  test/integration/test_integration.py::TestUtils::test_get_model_size_autoquant_5_cuda

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[cpuhrsch]

Another option here could be using a context.

[jerryzh168]

@cpuhrsch I'm planning to create another PR that is trying to subgraph extraction based on Vasiliy's code btw: https://github.com/pytorch-labs/ao_benchmarks/tree/main/experimental/float8_subgraph_analysis