[ROCm] Enable FP8 QAT test on MI300/MI350

[brucechanglongxu]

The FP8 QAT test (test_quantize_api_fp8_fp8) was gated on is_sm_at_least_89(), which excluded AMD GPUs. MI300 and MI350 both support FP8 dynamic activation + FP8 weight quantization with PerTensor and PerRow granularity, so this updates the skip condition to also allow MI300/MI350.

Tested locally on MI300X -- both PerTensor and PerRow variants pass.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4132

• 📄 Preview Python docs built from this PR

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

✅ You can merge normally! (1 Unrelated Failure)

As of commit 0d5a269 with merge base 4cef023 ():

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• Run Regression Tests / test-nightly (CUDA Nightly, linux.g5.12xlarge.nvidia.gpu, --pre torch --index-url https://downloa... / linux-job (gh) (trunk failure)
  test/test_low_bit_optim.py::TestFSDP2::test_fsdp2

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

[brucechanglongxu]

@andrewor14 Thanks for the approval! This is mergeable and CI is green (the CUDA nightly failure looks like a pre-existing infra issue). Could you merge when you get a chance?