`Trainer` with `logger=False` still produces logs

[ValerianRey]

Bug description

Hi! I need to maintain a test of a usage example of my library torchjd, showing how it can be combined with PyTorch Lightning. To keep my test outputs clean, I would like to disable all logs from PyTorch Lightning, but I can't manage to do so.

Even when passing logger=False to the Trainer, and when manually disabling logs, I can't get rid of the INFO: GPU available: [...] logs. I'm posting this as a bug since I think that passing logger=False is supposed to do precisely what I want, but maybe since a recent release it stopped working?

I only found outdated solutions on stackoverflow and in other issues. Thanks in advance for the help! :)

What version are you seeing the problem on?

v2.4

How to reproduce the bug

This reproducer is the first 4 steps of the tutorial, with 2 changes:

• The first 5 lines that try to disable logging manually
• The extra parameters enable_checkpointing=False, logger=False, enable_model_summary=False, enable_progress_bar=False provided to the Trainer

    import logging
    logging.getLogger("lightning_fabric").propagate = False
    logging.getLogger("lightning_fabric").setLevel = False
    logging.getLogger("pytorch_lightning").propagate = False
    logging.getLogger("pytorch_lightning").setLevel = False

    import os
    from torch import optim, nn, utils, Tensor
    from torchvision.datasets import MNIST
    from torchvision.transforms import ToTensor
    import lightning as L

    # define any number of nn.Modules (or use your current ones)
    encoder = nn.Sequential(nn.Linear(28 * 28, 64), nn.ReLU(), nn.Linear(64, 3))
    decoder = nn.Sequential(nn.Linear(3, 64), nn.ReLU(), nn.Linear(64, 28 * 28))

    # define the LightningModule
    class LitAutoEncoder(L.LightningModule):
        def __init__(self, encoder, decoder):
            super().__init__()
            self.encoder = encoder
            self.decoder = decoder

        def training_step(self, batch, batch_idx):
            # training_step defines the train loop.
            # it is independent of forward
            x, _ = batch
            x = x.view(x.size(0), -1)
            z = self.encoder(x)
            x_hat = self.decoder(z)
            loss = nn.functional.mse_loss(x_hat, x)
            # Logging to TensorBoard (if installed) by default
            self.log("train_loss", loss)
            return loss

        def configure_optimizers(self):
            optimizer = optim.Adam(self.parameters(), lr=1e-3)
            return optimizer

    # init the autoencoder
    autoencoder = LitAutoEncoder(encoder, decoder)

    # setup data
    dataset = MNIST(os.getcwd(), download=True, transform=ToTensor())
    train_loader = utils.data.DataLoader(dataset)

    # train the model (hint: here are some helpful Trainer arguments for rapid idea iteration)
    trainer = L.Trainer(limit_train_batches=100, max_epochs=1, enable_checkpointing=False, logger=False, enable_model_summary=False, enable_progress_bar=False)
    trainer.fit(model=autoencoder, train_dataloaders=train_loader)

Error messages and logs

Output (in red in my terminal):

INFO: GPU available: True (cuda), used: True
INFO: TPU available: False, using: 0 TPU cores
INFO: HPU available: False, using: 0 HPUs
INFO: LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
INFO: `Trainer.fit` stopped: `max_epochs=1` reached.

Environment

Current environment

• CUDA:
  - GPU:
  - NVIDIA GeForce GTX 1080
  - available: True
  - version: 12.4
• Lightning:
  - lightning: 2.4.0
  - lightning-utilities: 0.11.8
  - pytorch-lightning: 2.4.0
  - torch: 2.5.1
  - torchjd: 0.2.2
  - torchmetrics: 1.6.0
  - torchvision: 0.20.1
• Packages:
  - aiohappyeyeballs: 2.4.3
  - aiohttp: 3.11.2
  - aiosignal: 1.3.1
  - alabaster: 0.7.16
  - attrs: 24.2.0
  - autocommand: 2.2.2
  - babel: 2.16.0
  - backports.tarfile: 1.2.0
  - beautifulsoup4: 4.12.3
  - blinker: 1.9.0
  - certifi: 2024.8.30
  - cfgv: 3.4.0
  - charset-normalizer: 3.4.0
  - clarabel: 0.9.0
  - click: 8.1.7
  - coverage: 7.6.7
  - cvxpy: 1.6.0
  - dash: 2.18.2
  - dash-core-components: 2.0.0
  - dash-html-components: 2.0.0
  - dash-table: 5.0.0
  - distlib: 0.3.9
  - docutils: 0.21.2
  - ecos: 2.0.14
  - filelock: 3.16.1
  - flask: 3.0.3
  - frozenlist: 1.5.0
  - fsspec: 2024.10.0
  - furo: 2024.1.29
  - identify: 2.6.2
  - idna: 3.10
  - imagesize: 1.4.1
  - importlib-metadata: 8.5.0
  - inflect: 7.3.1
  - iniconfig: 2.0.0
  - itsdangerous: 2.2.0
  - jaraco.collections: 5.1.0
  - jaraco.context: 5.3.0
  - jaraco.functools: 4.0.1
  - jaraco.text: 3.12.1
  - jinja2: 3.1.4
  - kaleido: 0.2.1
  - lightning: 2.4.0
  - lightning-utilities: 0.11.8
  - markdown-it-py: 3.0.0
  - markupsafe: 3.0.2
  - mdit-py-plugins: 0.4.2
  - mdurl: 0.1.2
  - more-itertools: 10.3.0
  - mpmath: 1.3.0
  - multidict: 6.1.0
  - myst-parser: 4.0.0
  - nest-asyncio: 1.6.0
  - networkx: 3.4.2
  - nodeenv: 1.9.1
  - numpy: 2.1.3
  - nvidia-cublas-cu12: 12.4.5.8
  - nvidia-cuda-cupti-cu12: 12.4.127
  - nvidia-cuda-nvrtc-cu12: 12.4.127
  - nvidia-cuda-runtime-cu12: 12.4.127
  - nvidia-cudnn-cu12: 9.1.0.70
  - nvidia-cufft-cu12: 11.2.1.3
  - nvidia-curand-cu12: 10.3.5.147
  - nvidia-cusolver-cu12: 11.6.1.9
  - nvidia-cusparse-cu12: 12.3.1.170
  - nvidia-nccl-cu12: 2.21.5
  - nvidia-nvjitlink-cu12: 12.4.127
  - nvidia-nvtx-cu12: 12.4.127
  - osqp: 0.6.7.post3
  - packaging: 24.2
  - pillow: 11.0.0
  - platformdirs: 4.3.6
  - plotly: 5.24.1
  - pluggy: 1.5.0
  - pre-commit: 4.0.1
  - propcache: 0.2.0
  - pygments: 2.18.0
  - pytest: 8.3.3
  - pytest-cov: 6.0.0
  - pytorch-lightning: 2.4.0
  - pyyaml: 6.0.2
  - qdldl: 0.1.7.post4
  - qpsolvers: 4.4.0
  - quadprog: 0.1.13
  - requests: 2.32.3
  - retrying: 1.3.4
  - scipy: 1.14.1
  - scs: 3.2.7
  - setuptools: 75.6.0
  - six: 1.16.0
  - snowballstemmer: 2.2.0
  - soupsieve: 2.6
  - sphinx: 7.4.7
  - sphinx-autodoc-typehints: 2.3.0
  - sphinx-basic-ng: 1.0.0b2
  - sphinx-design: 0.6.1
  - sphinxcontrib-applehelp: 2.0.0
  - sphinxcontrib-devhelp: 2.0.0
  - sphinxcontrib-htmlhelp: 2.1.0
  - sphinxcontrib-jsmath: 1.0.1
  - sphinxcontrib-qthelp: 2.0.0
  - sphinxcontrib-serializinghtml: 2.0.0
  - sympy: 1.13.1
  - tenacity: 9.0.0
  - tomli: 2.1.0
  - torch: 2.5.1
  - torchjd: 0.2.2
  - torchmetrics: 1.6.0
  - torchvision: 0.20.1
  - tqdm: 4.67.0
  - triton: 3.1.0
  - typeguard: 4.3.0
  - typing-extensions: 4.12.2
  - urllib3: 2.2.3
  - virtualenv: 20.27.1
  - werkzeug: 3.0.6
  - wheel: 0.43.0
  - yarl: 1.17.1
  - zipp: 3.21.0
• System:
  - OS: Linux
  - architecture:
  - 64bit
  - ELF
  - processor: x86_64
  - python: 3.12.3
  - release: 5.15.0-126-generic
  - version: tensorboarX to tensorboardX #136~20.04.1-Ubuntu SMP Thu Nov 14 16:38:05 UTC 2024

More info

No response

[tshu-w]

As documented, the logger parameter in Trainer is specifically for controlling experiment tracking loggers (like Tensorboard), not Python's logging system.

For Python logging messages:

1. To disable INFO level logs, use Python's logging:

import logging
logging.disable(logging.INFO)

2. For warnings, please refer to the documentation on managing warnings:
   https://lightning.ai/docs/pytorch/stable/advanced/warnings.html

The logger=False parameter is working as intended - it only controls experiment tracking loggers, not python logging.

[ValerianRey]

Thanks, your solution works!