fix: mute unsupported field attribute warning on startup

[akoumpa]

What does this PR do ?

While confirming #1766 (comment) I saw there were more warning left, with this filtering

root@c16f65276617:/mnt/4tb/auto/26_04/Automodel# python3 app.py examples/llm_finetune/llama3_2/llama3_2_1b_squad_peft.yaml --nproc-per-node=2
INFO:__main__:Running from source checkout (app.py). For production use, install the package and run `automodel` or `am` instead.
INFO:nemo_automodel.cli.app:Config: /mnt/4tb/auto/26_04/Automodel/examples/llm_finetune/llama3_2/llama3_2_1b_squad_peft.yaml
INFO:nemo_automodel.cli.app:Recipe: nemo_automodel.recipes.llm.train_ft.``
INFO:nemo_automodel.cli.app:Launching job interactively (local)
cfg-path: /mnt/4tb/auto/26_04/Automodel/examples/llm_finetune/llama3_2/llama3_2_1b_squad_peft.yaml
/usr/local/lib/python3.12/dist-packages/torch/cuda/__init__.py:61: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report t
his to the maintainers of the package that installed pynvml for you.
  import pynvml  # type: ignore[import]
INFO:nemo_automodel.components.launcher.interactive:Running job using source from: /mnt/4tb/auto/26_04/Automodel
INFO:nemo_automodel.components.launcher.interactive:Launching job locally on 2 devices
W0410 19:01:12.990000 450 torch/distributed/run.py:851]
W0410 19:01:12.990000 450 torch/distributed/run.py:851] *****************************************
W0410 19:01:12.990000 450 torch/distributed/run.py:851] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variabl
e for optimal performance in your application as needed.
W0410 19:01:12.990000 450 torch/distributed/run.py:851] *****************************************
/usr/local/lib/python3.12/dist-packages/torch/cuda/__init__.py:61: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report t
his to the maintainers of the package that installed pynvml for you.
  import pynvml  # type: ignore[import]
/usr/local/lib/python3.12/dist-packages/torch/cuda/__init__.py:61: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report t
his to the maintainers of the package that installed pynvml for you.
  import pynvml  # type: ignore[import]
cfg-path: /mnt/4tb/auto/26_04/Automodel/examples/llm_finetune/llama3_2/llama3_2_1b_squad_peft.yaml
cfg-path: /mnt/4tb/auto/26_04/Automodel/examples/llm_finetune/llama3_2/llama3_2_1b_squad_peft.yaml
> initializing torch distributed with 2 workers.
2026-04-10 19:01:21 | INFO | nemo_automodel.components.loggers.log_utils | Setting logging level to 20
2026-04-10 19:01:21 | INFO | nemo_automodel.shared.te_patches | Applied FusedAdam QuantizedTensor monkey-patch.
2026-04-10 19:01:21 | INFO | root | Experiment_details:
2026-04-10 19:01:21 | INFO | root | Timestamp: '2026-04-10T19:01:21'
2026-04-10 19:01:21 | INFO | root | User: root
2026-04-10 19:01:21 | INFO | root | Host: c16f65276617
2026-04-10 19:01:21 | INFO | root | World size: 2
2026-04-10 19:01:21 | INFO | root | Backend: nccl
2026-04-10 19:01:21 | INFO | root | Recipe: TrainFinetuneRecipeForNextTokenPrediction
2026-04-10 19:01:21 | INFO | root | Model name: meta-llama/Llama-3.2-1B
recipe: TrainFinetuneRecipeForNextTokenPrediction
step_scheduler:
  global_batch_size: 64
  local_batch_size: 8
  ckpt_every_steps: 1000

Changelog

• Add specific line by line info of high level changes in this PR.

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation?

If you haven't finished some of the above items you can still open "Draft" PR.

Additional Information

• Related to # (issue)

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[akoumpa]

/ok to test 9673b6c