Merge branch 'master' into batch_size_scaler_edgecase

SkafteNicki · web-flow · commit e1cd01e9bc24 · 2025-09-09T07:25:50.000+02:00
diff --git a/.actions/assistant.py b/.actions/assistant.py
@@ -459,6 +459,25 @@ def generate_docker_tags(
         tags = [f"{docker_project}:{tag}" for tag in tags]
         print(",".join(tags))
 
+    @staticmethod
+    def prune_pytest_as_errors(
+        pyproject_toml: str = "pyproject.toml", errors: tuple = ("FutureWarning", "DeprecationWarning")
+    ) -> None:
+        """Prune pytest warnings as errors from the pyproject.toml file."""
+        import tomlkit
+
+        with open(pyproject_toml, encoding="utf-8") as fopen:
+            content = fopen.read()
+        pyproject = tomlkit.parse(content)
+        filterwarnings = pyproject.get("tool", {}).get("pytest", {}).get("ini_options", {}).get("filterwarnings", [])
+        if not filterwarnings:
+            return
+        filterwarnings = [wrn for wrn in filterwarnings if not any(f"error::{err}" in wrn for err in errors)]
+        pyproject["tool"]["pytest"]["ini_options"]["filterwarnings"] = filterwarnings
+
+        with open(pyproject_toml, "w", encoding="utf-8") as fopen:
+            fopen.write(tomlkit.dumps(pyproject))
+
 
 if __name__ == "__main__":
     import jsonargparse
diff --git a/.github/workflows/ci-tests-fabric.yml b/.github/workflows/ci-tests-fabric.yml
@@ -99,11 +99,14 @@ jobs:
       - name: Set min. dependencies
         if: ${{ matrix.config.requires == 'oldest' }}
         run: |
-          cd requirements/fabric
           uv pip install -U "lightning-utilities[cli]"
-          python -m lightning_utilities.cli requirements set-oldest --req_files "['base.txt', 'strategies.txt', 'test.txt']"
+          python -m lightning_utilities.cli requirements set-oldest \
+            --req_files "['requirements/fabric/base.txt', 'requirements/fabric/strategies.txt', 'requirements/fabric/test.txt']"
           uv pip install "cython<3.0" wheel
           uv pip install "pyyaml==5.4" --no-build-isolation
+          # This script removes any line containing "error::FutureWarning" from pyproject.toml
+          uv pip install -r requirements/ci.txt
+          python .actions/assistant.py prune_pytest_as_errors
 
       - name: Adjust PyTorch versions in requirements files
         if: ${{ matrix.config.requires != 'oldest' }}
diff --git a/.github/workflows/ci-tests-pytorch.yml b/.github/workflows/ci-tests-pytorch.yml
@@ -98,11 +98,14 @@ jobs:
       - name: Set min. dependencies
         if: ${{ matrix.config.requires == 'oldest' }}
         run: |
-          cd requirements/pytorch
           uv pip install -U "lightning-utilities[cli]"
-          python -m lightning_utilities.cli requirements set-oldest --req_files "['base.txt', 'extra.txt', 'strategies.txt', 'examples.txt', 'test.txt']"
+          python -m lightning_utilities.cli requirements set-oldest \
+            --req_files "['requirements/pytorch/base.txt', 'requirements/pytorch/extra.txt', 'requirements/pytorch/strategies.txt', 'requirements/pytorch/examples.txt', 'requirements/pytorch/test.txt']"
           uv pip install "cython<3.0" wheel
           uv pip install "pyyaml==5.4" --no-build-isolation
+          # This script removes any line containing "error::FutureWarning" from pyproject.toml
+          uv pip install -r requirements/ci.txt
+          python .actions/assistant.py prune_pytest_as_errors
 
       - name: Adjust PyTorch versions in requirements files
         if: ${{ matrix.config.requires != 'oldest' }}
diff --git a/.github/workflows/code-checks.yml b/.github/workflows/code-checks.yml
@@ -29,23 +29,32 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - uses: actions/checkout@v5
-      - uses: actions/setup-python@v6
-        with:
-          python-version: "3.11"
 
-      - name: Mypy cache
-        uses: actions/cache@v4
+      - name: Install uv and set Python version
+        uses: astral-sh/setup-uv@v6
         with:
-          path: .mypy_cache
-          key: mypy-${{ hashFiles('requirements/typing.txt') }}
+          python-version: "3.11"
+          # TODO: Avoid activating environment like this
+          # see: https://github.com/astral-sh/setup-uv/tree/v6/?tab=readme-ov-file#activate-environment
+          activate-environment: true
+          enable-cache: true
 
       - name: Install dependencies
         env:
           FREEZE_REQUIREMENTS: 1
         timeout-minutes: 20
         run: |
-          pip install -e '.[pytorch-all,fabric-all]' -r requirements/typing.txt
-          pip list
+          uv pip install '.[pytorch-all,fabric-all]' -r requirements/typing.txt
+          uv pip list
+
+      - name: mypy cache
+        uses: actions/cache@v4
+        with:
+          path: .mypy_cache
+          key: mypy-${{ hashFiles('requirements/typing.txt') }}
 
       - name: Check typing
         run: mypy
+
+      - name: Minimize uv cache
+        run: uv cache prune --ci
diff --git a/pyproject.toml b/pyproject.toml
@@ -182,7 +182,6 @@ filterwarnings = [
     # "error::DeprecationWarning",
     "error::FutureWarning",
     "ignore::FutureWarning:onnxscript", # Temporary ignore until onnxscript is updated
-    "ignore:The pynvml package is deprecated:FutureWarning", # Ignore pynvml deprecation warning, since it is not installed by PL directly
 ]
 xfail_strict = true
 junit_duration_report = "call"
diff --git a/requirements/ci.txt b/requirements/ci.txt
@@ -6,3 +6,4 @@ importlib-metadata <9.0.0
 wget
 pkginfo ==1.12.1.2
 packaging <25.1
+tomlkit
diff --git a/requirements/fabric/test.txt b/requirements/fabric/test.txt
@@ -1,7 +1,7 @@
 coverage ==7.10.6
 numpy >=1.21.0, <1.27.0
 pytest ==8.4.2
-pytest-cov ==6.2.1
+pytest-cov ==6.3.0
 pytest-timeout ==2.4.0
 pytest-rerunfailures ==16.0.1
 pytest-random-order ==1.2.0
diff --git a/requirements/pytorch/examples.txt b/requirements/pytorch/examples.txt
@@ -3,5 +3,5 @@
 
 requests <2.33.0
 torchvision >=0.16.0, <0.24.0
-ipython[all] >=8.0.0, <9.0.0
+ipython[all] >=8.0.0, <10.0.0
 torchmetrics >=0.10.0, <1.9.0
diff --git a/requirements/pytorch/test.txt b/requirements/pytorch/test.txt
@@ -1,6 +1,6 @@
 coverage ==7.10.6
 pytest ==8.4.2
-pytest-cov ==6.2.1
+pytest-cov ==6.3.0
 pytest-timeout ==2.4.0
 pytest-rerunfailures ==16.0.1
 pytest-random-order ==1.2.0
diff --git a/src/lightning/fabric/CHANGELOG.md b/src/lightning/fabric/CHANGELOG.md
@@ -9,9 +9,6 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Added
 
-- Added `exclude_frozen_parameters` to `DeepSpeedStrategy` ([#21060](https://github.com/Lightning-AI/pytorch-lightning/pull/21060))
-
-
 -
 
 
@@ -22,20 +19,33 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Changed
 
-- let `_get_default_process_group_backend_for_device` support more hardware platforms (
-    [#21057](https://github.com/Lightning-AI/pytorch-lightning/pull/21057), [#21093](https://github.com/Lightning-AI/pytorch-lightning/pull/21093))
+-
 
 
-### Fixed
+- Set `_DeviceDtypeModuleMixin._device` from torch's default device function ([#21164](https://github.com/Lightning-AI/pytorch-lightning/pull/21164))
 
-- Fixed with adding a missing device id for pytorch 2.8 ([#21105](https://github.com/Lightning-AI/pytorch-lightning/pull/21105))
 
+### Fixed
 
-- Respect `verbose=False` in `seed_everything` when no seed is provided
+-
 
 
 ---
 
+## [2.5.5] - 2025-09-05
+
+### Changed
+
+- Include `exclude_frozen_parameters` to `DeepSpeedStrategy` ([#21060](https://github.com/Lightning-AI/pytorch-lightning/pull/21060))
+- Let `_get_default_process_group_backend_for_device` support more hardware platforms (
+    [#21057](https://github.com/Lightning-AI/pytorch-lightning/pull/21057), [#21093](https://github.com/Lightning-AI/pytorch-lightning/pull/21093))
+
+### Fixed
+
+- Fixed with adding a missing device id for pytorch 2.8 ([#21105](https://github.com/Lightning-AI/pytorch-lightning/pull/21105))
+- Respecting `verbose=False` in `seed_everything` when no seed is provided ([#21161](https://github.com/Lightning-AI/pytorch-lightning/pull/21161))
+
+
 ## [2.5.4] - 2025-08-29
 
 ### Changed
diff --git a/src/lightning/fabric/utilities/device_dtype_mixin.py b/src/lightning/fabric/utilities/device_dtype_mixin.py
@@ -18,14 +18,18 @@
 from torch.nn import Module
 from typing_extensions import Self, override
 
+from lightning.fabric.utilities.imports import _TORCH_GREATER_EQUAL_2_3
+
 
 class _DeviceDtypeModuleMixin(Module):
     __jit_unused_properties__: list[str] = ["device", "dtype"]
 
     def __init__(self) -> None:
         super().__init__()
         self._dtype: Union[str, torch.dtype] = torch.get_default_dtype()
-        self._device = torch.device("cpu")
+        # Workarounds from the original pytorch issue:
+        # https://github.com/pytorch/pytorch/issues/115333#issuecomment-1848449687
+        self._device = torch.get_default_device() if _TORCH_GREATER_EQUAL_2_3 else torch.empty(0).device
 
     @property
     def dtype(self) -> Union[str, torch.dtype]:
diff --git a/src/lightning/pytorch/CHANGELOG.md b/src/lightning/pytorch/CHANGELOG.md
@@ -16,13 +16,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added Torch-Tensorrt integration with `LightningModule` ([#20808](https://github.com/Lightning-AI/pytorch-lightning/pull/20808))
 
 
-- Added `exclude_frozen_parameters` to `DeepSpeedStrategy` ([#21060](https://github.com/Lightning-AI/pytorch-lightning/pull/21060))
-
-
-- Added `PossibleUserWarning` that is raised if modules are in eval mode when training starts ([#21146](https://github.com/Lightning-AI/pytorch-lightning/pull/21146))
-
-
-- Added time based validation support though `val_check_interval` ([#21071](https://github.com/Lightning-AI/pytorch-lightning/pull/21071))
+- Added time-based validation support though `val_check_interval` ([#21071](https://github.com/Lightning-AI/pytorch-lightning/pull/21071))
 
 
 ### Changed
@@ -37,25 +31,27 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Fixed
 
-- Fixed `LightningCLI` not using `ckpt_path` hyperparameters to instantiate classes ([#21116](https://github.com/Lightning-AI/pytorch-lightning/pull/21116))
+- Fixed edgecase when `max_trials` is reached in `Tuner.scale_batch_size` ([#21187](https://github.com/Lightning-AI/pytorch-lightning/pull/21187))
 
 
-- Fixed callbacks by defer step/time-triggered `ModelCheckpoint` saves until validation metrics are available ([#21106](https://github.com/Lightning-AI/pytorch-lightning/pull/21106))
+---
 
+## [2.5.5] - 2025-09-05
 
-- Fixed with adding a missing device id for pytorch 2.8 ([#21105](https://github.com/Lightning-AI/pytorch-lightning/pull/21105))
+### Changed
 
+- Include `exclude_frozen_parameters` to `DeepSpeedStrategy` ([#21060](https://github.com/Lightning-AI/pytorch-lightning/pull/21060))
+- Include `PossibleUserWarning` that is raised if modules are in eval mode when training starts ([#21146](https://github.com/Lightning-AI/pytorch-lightning/pull/21146))
 
-- Fixed `TQDMProgressBar` not resetting correctly when using both a finite and iterable dataloader ([#21147](https://github.com/Lightning-AI/pytorch-lightning/pull/21147))
-
+### Fixed
 
+- Fixed `LightningCLI` not using `ckpt_path` hyperparameters to instantiate classes ([#21116](https://github.com/Lightning-AI/pytorch-lightning/pull/21116))
+- Fixed callbacks by defer step/time-triggered `ModelCheckpoint` saves until validation metrics are available ([#21106](https://github.com/Lightning-AI/pytorch-lightning/pull/21106))
+- Fixed with adding a missing device id for pytorch 2.8 ([#21105](https://github.com/Lightning-AI/pytorch-lightning/pull/21105))
+- Fixed `TQDMProgressBar` not resetting correctly when using both a finite and iterable dataloader ([#21147](https://github.com/Lightning-AI/pytorch-lightning/pull/21147))
 - Fixed cleanup of temporary files from `Tuner` on crashes ([#21162](https://github.com/Lightning-AI/pytorch-lightning/pull/21162))
 
 
-- Fixed edgecase when `max_trials` is reached in `Tuner.scale_batch_size` ([#21187](https://github.com/Lightning-AI/pytorch-lightning/pull/21187))
-
----
-
 ## [2.5.4] - 2025-08-29
 
 ### Fixed
diff --git a/tests/tests_fabric/utilities/test_device_dtype_mixin.py b/tests/tests_fabric/utilities/test_device_dtype_mixin.py
@@ -2,6 +2,7 @@
 import torch
 from torch import nn as nn
 
+from lightning.fabric.plugins.precision.utils import _DtypeContextManager
 from lightning.fabric.utilities.device_dtype_mixin import _DeviceDtypeModuleMixin
 from tests_fabric.helpers.runif import RunIf
 
@@ -50,6 +51,30 @@ def test_submodules_device_and_dtype(dst_device_str, dst_type):
     assert model.dtype == model.module.module.dtype == dst_type
 
 
+@pytest.mark.parametrize(
+    "dst_device_str",
+    [
+        "cpu",
+        pytest.param("cuda:0", marks=RunIf(min_cuda_gpus=1)),
+        pytest.param("mps:0", marks=RunIf(mps=True)),
+    ],
+)
+@pytest.mark.parametrize(
+    "dst_type",
+    [
+        torch.float,
+        pytest.param(torch.half, marks=RunIf(mps=False)),
+        pytest.param(torch.double, marks=RunIf(mps=False)),
+    ],
+)
+def test_submodules_context_device_and_dtype(dst_device_str, dst_type):
+    dst_device = torch.device(dst_device_str)
+    with _DtypeContextManager(dst_type), dst_device:
+        model = TopModule()
+    assert model.device == dst_device
+    assert model.dtype == dst_type
+
+
 @pytest.mark.parametrize(
     "device",
     [
diff --git a/tests/tests_pytorch/trainer/test_trainer.py b/tests/tests_pytorch/trainer/test_trainer.py
@@ -2107,6 +2107,22 @@ def test_init_module_context(monkeypatch):
     strategy.tensor_init_context.reset_mock()
 
 
+@pytest.mark.parametrize(
+    ("target_device", "accelerator", "devices"),
+    [
+        ("cpu", "cpu", "auto"),
+        pytest.param("cuda:0", "gpu", [0], marks=RunIf(min_cuda_gpus=1)),
+        pytest.param("cuda:1", "gpu", [1], marks=RunIf(min_cuda_gpus=2)),
+    ],
+)
+def test_init_module_device_type(target_device, accelerator, devices):
+    """Test that the strategy returns the context manager for initializing the module."""
+    trainer = Trainer(accelerator=accelerator, devices=devices)
+    with trainer.init_module():
+        model = BoringModel()
+        assert model.device == torch.device(target_device)
+
+
 def test_expand_home_trainer():
     """Test that the dirpath gets expanded if it contains `~`."""
     home_root = Path.home()

Original file line number	Diff line number	Diff line change
`@@ -182,7 +182,6 @@ filterwarnings = [`
`182`	`182`	`# "error::DeprecationWarning",`
`183`	`183`	`"error::FutureWarning",`
`184`	`184`	`"ignore::FutureWarning:onnxscript", # Temporary ignore until onnxscript is updated`
`185`		`- "ignore:The pynvml package is deprecated:FutureWarning", # Ignore pynvml deprecation warning, since it is not installed by PL directly`
`186`	`185`	`]`
`187`	`186`	`xfail_strict = true`
`188`	`187`	`junit_duration_report = "call"`