Qualcomm AI Engine Direct - tensor dumping at lpai backend

[jethroqti]

Summary

Currently, using --dump_intermediate_outputs and RPC mode will dump tensor

Test plan

python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNQuantizedUtils.test_qnn_backend_dump_intermediate_outputs_simple_model -b build-android -s ${SN} -m ${CHIPID} --backend lpai python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNQuantizedUtils.test_qnn_backend_dump_intermediate_outputs_simple_model -b build-android -s ${SN} -m ${CHIPID} --dump_intermediate_outputs --backend lpai

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20369

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• [ROCm] MI350 CI jobs will have longer queue times due to CI migration

❌ 3 New Failures, 4 Pending, 1 Unrelated Failure

As of commit 4dc27a2 with merge base ef5c8a7 ():

NEW FAILURES - The following jobs have failed:

• pull / test-arm-backend-no-driver (test_pytest_models_tosa) / linux-job (gh)
  RuntimeError: Command docker exec -t 574efb08f1505fa3273d59a509837670cc029b1bd7e9c010160cf4e22e9ded31 /exec failed with exit code 1
• pull / test-lora-linux / linux-job (gh)
  RuntimeError: Command docker exec -t a942ddf247233cd217858052e2ffb2f116d5545a015e8515cd600909e1cfbe8d /exec failed with exit code 1
• pull / test-phi-3-mini-runner-linux / linux-job (gh)
  RuntimeError: Command docker exec -t 4ddcce6ed944e692cf92c22899a636a7d0309fc5d66cccbce757f4f7d37e64fb /exec failed with exit code 1

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

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

• pull / test-llama-runner-qnn-linux (fp32, qnn_16a16w, qnn) / linux-job (gh) (trunk failure)

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

[jethroqti]

@pytorchbot label "release notes: qualcomm"

[jethroqti]

@psiddh
This PR is one kind of tensor dumping at LPAI backend for debugging.
If you are available, please take a look. Thanks.
@shewu-quic