Merge branch 'master' into init_module_deepspeed_docs_fix

Borda · web-flow · commit 9f75aac732c4 · 2025-04-03T12:49:55.000-04:00
diff --git a/.azure/gpu-tests-fabric.yml b/.azure/gpu-tests-fabric.yml
@@ -62,6 +62,9 @@ jobs:
         "Fabric | latest":
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.6-cuda12.4.1"
           PACKAGE_NAME: "fabric"
+        "Fabric | future":
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.7-cuda12.6.3"
+          PACKAGE_NAME: "fabric"
         "Lightning | latest":
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.6-cuda12.4.1"
           PACKAGE_NAME: "lightning"
diff --git a/.azure/gpu-tests-pytorch.yml b/.azure/gpu-tests-pytorch.yml
@@ -55,6 +55,9 @@ jobs:
         "PyTorch | latest":
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.6-cuda12.4.1"
           PACKAGE_NAME: "pytorch"
+        "PyTorch | future":
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.7-cuda12.6.3"
+          PACKAGE_NAME: "pytorch"
         "Lightning | latest":
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.6-cuda12.4.1"
           PACKAGE_NAME: "lightning"
diff --git a/.github/checkgroup.yml b/.github/checkgroup.yml
@@ -127,12 +127,14 @@ subprojects:
       - "build-cuda (3.11, 2.3.1, 12.1.1)"
       - "build-cuda (3.11, 2.4.1, 12.1.1)"
       - "build-cuda (3.12, 2.5.1, 12.1.1)"
+      - "build-cuda (3.12, 2.6.0, 12.4.1)"
       #- "build-NGC"
       - "build-pl (3.10, 2.1, 12.1.1)"
       - "build-pl (3.11, 2.2, 12.1.1)"
       - "build-pl (3.11, 2.3, 12.1.1)"
       - "build-pl (3.11, 2.4, 12.1.1)"
-      - "build-pl (3.12, 2.5, 12.1.1, true)"
+      - "build-pl (3.12, 2.5, 12.1.1)"
+      - "build-pl (3.12, 2.6, 12.4.1, true)"
 
   # SECTION: lightning_fabric
 
diff --git a/.github/workflows/_legacy-checkpoints.yml b/.github/workflows/_legacy-checkpoints.yml
@@ -144,7 +144,7 @@ jobs:
           title: Adding test for legacy checkpoint created with ${{ env.PL_VERSION }}
           committer: GitHub <noreply@github.com>
           author: ${{ github.actor }} <${{ github.actor }}@users.noreply.github.com>
-          commit-message: "update tutorials to `${{ env.PL_VERSION }}`"
+          commit-message: "adding `${{ env.PL_VERSION }}` checkpoint"
           body: "**This is automated addition of created checkpoints with the latest `lightning` release!**"
           delete-branch: true
           token: ${{ secrets.PAT_GHOST }}
diff --git a/.github/workflows/call-clear-cache.yml b/.github/workflows/call-clear-cache.yml
@@ -23,18 +23,18 @@ on:
 jobs:
   cron-clear:
     if: github.event_name == 'schedule' || github.event_name == 'pull_request'
-    uses: Lightning-AI/utilities/.github/workflows/cleanup-caches.yml@v0.14.0
+    uses: Lightning-AI/utilities/.github/workflows/cleanup-caches.yml@v0.14.2
     with:
-      scripts-ref: v0.14.0
+      scripts-ref: v0.14.2
       dry-run: ${{ github.event_name == 'pull_request' }}
       pattern: "latest|docs"
       age-days: 7
 
   direct-clear:
     if: github.event_name == 'workflow_dispatch' || github.event_name == 'pull_request'
-    uses: Lightning-AI/utilities/.github/workflows/cleanup-caches.yml@v0.14.0
+    uses: Lightning-AI/utilities/.github/workflows/cleanup-caches.yml@v0.14.2
     with:
-      scripts-ref: v0.14.0
+      scripts-ref: v0.14.2
       dry-run: ${{ github.event_name == 'pull_request' }}
       pattern: ${{ inputs.pattern || 'pypi_wheels' }} # setting str in case of PR / debugging
       age-days: ${{ fromJSON(inputs.age-days) || 0 }} # setting 0 in case of PR / debugging
diff --git a/.github/workflows/ci-check-md-links.yml b/.github/workflows/ci-check-md-links.yml
@@ -14,7 +14,7 @@ on:
 
 jobs:
   check-md-links:
-    uses: Lightning-AI/utilities/.github/workflows/check-md-links.yml@v0.14.0
+    uses: Lightning-AI/utilities/.github/workflows/check-md-links.yml@v0.14.2
     with:
       config-file: ".github/markdown-links-config.json"
       base-branch: "master"
diff --git a/.github/workflows/ci-schema.yml b/.github/workflows/ci-schema.yml
@@ -8,7 +8,7 @@ on:
 
 jobs:
   check:
-    uses: Lightning-AI/utilities/.github/workflows/check-schema.yml@v0.14.0
+    uses: Lightning-AI/utilities/.github/workflows/check-schema.yml@v0.14.2
     with:
       # skip azure due to the wrong schema file by MSFT
       # https://github.com/Lightning-AI/lightning-flash/pull/1455#issuecomment-1244793607
diff --git a/.github/workflows/ci-tests-fabric.yml b/.github/workflows/ci-tests-fabric.yml
@@ -68,9 +68,9 @@ jobs:
           - { os: "ubuntu-20.04", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.5" }
           - { os: "windows-2022", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.5" }
           # adding recently cut Torch 2.7 - FUTURE
-          #  - { os: "macOS-14", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
-          #  - { os: "ubuntu-22.04", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
-          #  - { os: "windows-2022", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "macOS-14", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "ubuntu-22.04", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "windows-2022", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.7" }
     timeout-minutes: 25 # because of building grpcio on Mac
     env:
       PACKAGE_NAME: ${{ matrix.pkg-name }}
diff --git a/.github/workflows/ci-tests-pytorch.yml b/.github/workflows/ci-tests-pytorch.yml
@@ -72,9 +72,9 @@ jobs:
           - { os: "ubuntu-20.04", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.5" }
           - { os: "windows-2022", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.5" }
           # adding recently cut Torch 2.7 - FUTURE
-          #  - { os: "macOS-14", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
-          #  - { os: "ubuntu-22.04", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
-          #  - { os: "windows-2022", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "macOS-14", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "ubuntu-22.04", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
+          - { os: "windows-2022", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.7" }
     timeout-minutes: 50
     env:
       PACKAGE_NAME: ${{ matrix.pkg-name }}
diff --git a/.github/workflows/docker-build.yml b/.github/workflows/docker-build.yml
@@ -47,7 +47,8 @@ jobs:
           - { python_version: "3.11", pytorch_version: "2.2", cuda_version: "12.1.1" }
           - { python_version: "3.11", pytorch_version: "2.3", cuda_version: "12.1.1" }
           - { python_version: "3.11", pytorch_version: "2.4", cuda_version: "12.1.1" }
-          - { python_version: "3.12", pytorch_version: "2.5", cuda_version: "12.1.1", latest: "true" }
+          - { python_version: "3.12", pytorch_version: "2.5", cuda_version: "12.1.1" }
+          - { python_version: "3.12", pytorch_version: "2.6", cuda_version: "12.4.1", latest: "true" }
     steps:
       - uses: actions/checkout@v4
         with:
@@ -101,6 +102,7 @@ jobs:
           - { python_version: "3.11", pytorch_version: "2.4.1", cuda_version: "12.1.1" }
           - { python_version: "3.12", pytorch_version: "2.5.1", cuda_version: "12.1.1" }
           - { python_version: "3.12", pytorch_version: "2.6.0", cuda_version: "12.4.1" }
+          - { python_version: "3.12", pytorch_version: "2.7.0", cuda_version: "12.6.3" }
     steps:
       - uses: actions/checkout@v4
       - uses: docker/setup-buildx-action@v3
@@ -131,7 +133,7 @@ jobs:
           status: ${{ job.status }}
           token: ${{ secrets.GITHUB_TOKEN }}
           notification_title: ${{ format('CUDA; {0} py{1} for *{2}*', runner.os, matrix.python_version, matrix.pytorch_version) }}
-          message_format: "{emoji} *{workflow}* {status_message}, see <{run_url}|detail>, cc: <@U01A5T7EY9M>" # akihironitta
+          message_format: "{emoji} *{workflow}* {status_message}, see <{run_url}|detail>, cc: <@U01A5T7EY9M>"
         env:
           SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}
 
diff --git a/Makefile b/Makefile
@@ -53,7 +53,7 @@ docs-fabric: clean sphinx-theme
 	cd docs/source-fabric && $(MAKE) html --jobs $(nproc)
 
 docs-pytorch: clean sphinx-theme
-	pip install -e .[all] --quiet -r requirements/pytorch/docs.txt -r _notebooks/.actions/requires.txt
+	pip install -e .[all] --quiet -r requirements/pytorch/docs.txt
 	cd docs/source-pytorch && $(MAKE) html --jobs $(nproc)
 
 update:
diff --git a/requirements/fabric/test.txt b/requirements/fabric/test.txt
@@ -7,4 +7,3 @@ pytest-rerunfailures ==12.0
 pytest-random-order ==1.1.0
 click ==8.1.7
 tensorboardX >=2.2, <2.7.0  # min version is set by torch.onnx missing attribute
-torchmetrics >=0.7.0, <1.5.0 # needed for using fixed compare_version
diff --git a/requirements/pytorch/base.txt b/requirements/pytorch/base.txt
@@ -5,7 +5,7 @@ torch >=2.1.0, <2.6.0
 tqdm >=4.57.0, <4.67.0
 PyYAML >=5.4, <6.1.0
 fsspec[http] >=2022.5.0, <2024.4.0
-torchmetrics >=0.7.0, <1.5.0  # needed for using fixed compare_version
+torchmetrics >=0.7.0, <1.5.0
 packaging >=20.0, <=23.1
 typing-extensions >=4.4.0, <4.11.0
 lightning-utilities >=0.10.0, <0.12.0
diff --git a/requirements/pytorch/docs.txt b/requirements/pytorch/docs.txt
@@ -4,4 +4,4 @@ nbformat  # used for generate empty notebook
 ipython[notebook] <8.7.0
 setuptools<58.0  # workaround for `error in ipython setup command: use_2to3 is invalid.`
 
--r ../../_notebooks/.actions/requires.txt
+#-r ../../_notebooks/.actions/requires.txt
diff --git a/src/lightning/fabric/CHANGELOG.md b/src/lightning/fabric/CHANGELOG.md
@@ -4,16 +4,30 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
+
 ## [unreleased] - YYYY-MM-DD
 
 ### Added
 
-- Added logging support for list of dicts without collapsing to a single key ([#19957](https://github.com/Lightning-AI/pytorch-lightning/issues/19957))
+-
+
+
+### Removed
+
+-
 
 
+---
+
+## [2.5.1] - 2025-03-18
+
+### Changed
+
+- Added logging support for list of dicts without collapsing to a single key ([#19957](https://github.com/Lightning-AI/pytorch-lightning/issues/19957))
+
 ### Removed
 
-- Removed legacy supoport for `lightning run model`. Use `fabric run` instead. ([#20588](https://github.com/Lightning-AI/pytorch-lightning/pull/20588))
+- Removed legacy support for `lightning run model`. Use `fabric run` instead. ([#20588](https://github.com/Lightning-AI/pytorch-lightning/pull/20588))
 
 
 ## [2.5.0] - 2024-12-19
diff --git a/src/lightning/fabric/accelerators/registry.py b/src/lightning/fabric/accelerators/registry.py
@@ -107,9 +107,9 @@ def remove(self, name: str) -> None:
         """Removes the registered accelerator by name."""
         self.pop(name)
 
-    def available_accelerators(self) -> list[str]:
-        """Returns a list of registered accelerators."""
-        return list(self.keys())
+    def available_accelerators(self) -> set[str]:
+        """Returns a set of registered accelerators."""
+        return set(self.keys())
 
     def __str__(self) -> str:
         return "Registered Accelerators: {}".format(", ".join(self.available_accelerators()))
diff --git a/src/lightning/fabric/utilities/spike.py b/src/lightning/fabric/utilities/spike.py
@@ -52,7 +52,7 @@ def __init__(
             from torchmetrics.aggregation import MeanMetric
             from torchmetrics.wrappers import Running
         else:
-            raise RuntimeError("SpikeDetection requires torchmetrics>=1.0.0! Please upgrade your version!")
+            raise RuntimeError("SpikeDetection requires `torchmetrics>=1.0.0` Please upgrade your version.")
         super().__init__()
 
         self.last_val: Union[torch.Tensor, float] = 0.0
diff --git a/src/lightning/pytorch/CHANGELOG.md b/src/lightning/pytorch/CHANGELOG.md
@@ -4,23 +4,17 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
+
 ## [unreleased] - YYYY-MM-DD
 
 ### Added
 
 - Add enable_autolog_hparams argument to Trainer ([#20593](https://github.com/Lightning-AI/pytorch-lightning/pull/20593))
 
 
-- Allow LightningCLI to use a customized argument parser class ([#20596](https://github.com/Lightning-AI/pytorch-lightning/pull/20596))
-
-
 ### Changed
 
-- Change `wandb` default x-axis to `tensorboard`'s `global_step` when `sync_tensorboard=True` ([#20611](https://github.com/Lightning-AI/pytorch-lightning/pull/20611))
-
-
-- Added a new `checkpoint_path_prefix` parameter to the MLflow logger which can control the path to where the MLflow artifacts for the model checkpoints are stored ([#20538](https://github.com/Lightning-AI/pytorch-lightning/pull/20538))
-- CometML logger was updated to support the recent Comet SDK ([#20275](https://github.com/Lightning-AI/pytorch-lightning/pull/20275))
+-
 
 
 ### Removed
@@ -30,12 +24,33 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Fixed
 
-- Fix CSVLogger logging hyperparameter at every write which increase latency  ([#20594](https://github.com/Lightning-AI/pytorch-lightning/pull/20594))
+-
 
-- Fix OverflowError when resuming from checkpoint with an iterable dataset ([#20565](https://github.com/Lightning-AI/pytorch-lightning/issues/20565))
 
+---
+
+## [2.5.1] - 2025-03-18
+
+### Changed
+
+- Allow LightningCLI to use a customized argument parser class ([#20596](https://github.com/Lightning-AI/pytorch-lightning/pull/20596))
+- Change `wandb` default x-axis to `tensorboard`'s `global_step` when `sync_tensorboard=True` ([#20611](https://github.com/Lightning-AI/pytorch-lightning/pull/20611))
+- Added a new `checkpoint_path_prefix` parameter to the MLflow logger which can control the path to where the MLflow artifacts for the model checkpoints are stored ([#20538](https://github.com/Lightning-AI/pytorch-lightning/pull/20538))
+- CometML logger was updated to support the recent Comet SDK ([#20275](https://github.com/Lightning-AI/pytorch-lightning/pull/20275))
+- bump: testing with latest `torch` 2.6 ([#20509](https://github.com/Lightning-AI/pytorch-lightning/pull/20509))
+
+### Fixed
 
+- Fixed CSVLogger logging hyperparameter at every write which increase latency  ([#20594](https://github.com/Lightning-AI/pytorch-lightning/pull/20594))
+- Fixed OverflowError when resuming from checkpoint with an iterable dataset ([#20565](https://github.com/Lightning-AI/pytorch-lightning/issues/20565))
+- Fixed swapped _R_co and _P to prevent type error ([#20508](https://github.com/Lightning-AI/pytorch-lightning/issues/20508))
 - Always call `WandbLogger.experiment` first in `_call_setup_hook` to ensure `tensorboard` logs can sync to `wandb` ([#20610](https://github.com/Lightning-AI/pytorch-lightning/pull/20610))
+- Fixed TBPTT example ([#20528](https://github.com/Lightning-AI/pytorch-lightning/pull/20528))
+- Fixed test compatibility as AdamW became subclass of Adam ([#20574](https://github.com/Lightning-AI/pytorch-lightning/pull/20574))
+- Fixed file extension of model checkpoints uploaded by NeptuneLogger ([#20581](https://github.com/Lightning-AI/pytorch-lightning/pull/20581))
+- Reset trainer variable `should_stop` when `fit` is called ([#19177](https://github.com/Lightning-AI/pytorch-lightning/pull/19177))
+- Fixed making `WandbLogger` upload models from all `ModelCheckpoint` callbacks, not just one ([#20191](https://github.com/Lightning-AI/pytorch-lightning/pull/20191))
+- Error when logging to MLFlow deleted experiment ([#20556](https://github.com/Lightning-AI/pytorch-lightning/pull/20556))
 
 
 ## [2.5.0] - 2024-12-19
diff --git a/src/lightning/pytorch/loggers/mlflow.py b/src/lightning/pytorch/loggers/mlflow.py
@@ -180,7 +180,7 @@ def experiment(self) -> "MlflowClient":
 
         if self._experiment_id is None:
             expt = self._mlflow_client.get_experiment_by_name(self._experiment_name)
-            if expt is not None:
+            if expt is not None and expt.lifecycle_stage != "deleted":
                 self._experiment_id = expt.experiment_id
             else:
                 log.warning(f"Experiment with name {self._experiment_name} not found. Creating it.")
@@ -363,13 +363,13 @@ def _scan_and_log_checkpoints(self, checkpoint_callback: ModelCheckpoint) -> Non
             aliases = ["latest", "best"] if p == checkpoint_callback.best_model_path else ["latest"]
 
             # Artifact path on mlflow
-            artifact_path = Path(self._checkpoint_path_prefix) / Path(p).stem
+            artifact_path = Path(self._checkpoint_path_prefix, Path(p).stem).as_posix()
 
             # Log the checkpoint
             self.experiment.log_artifact(self._run_id, p, artifact_path)
 
             # Create a temporary directory to log on mlflow
-            with tempfile.TemporaryDirectory(prefix="test", suffix="test", dir=os.getcwd()) as tmp_dir:
+            with tempfile.TemporaryDirectory() as tmp_dir:
                 # Log the metadata
                 with open(f"{tmp_dir}/metadata.yaml", "w") as tmp_file_metadata:
                     yaml.dump(metadata, tmp_file_metadata, default_flow_style=False)
diff --git a/src/lightning/pytorch/trainer/__init__.py b/src/lightning/pytorch/trainer/__init__.py
@@ -11,7 +11,6 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-""""""
 
 from lightning.fabric.utilities.seed import seed_everything
 from lightning.pytorch.trainer.trainer import Trainer
diff --git a/src/lightning/pytorch/trainer/connectors/callback_connector.py b/src/lightning/pytorch/trainer/connectors/callback_connector.py
@@ -11,13 +11,15 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-
+import inspect
 import logging
 import os
 from collections.abc import Sequence
 from datetime import timedelta
 from typing import Optional, Union
 
+from lightning_utilities import module_available
+
 import lightning.pytorch as pl
 from lightning.fabric.utilities.registry import _load_external_callbacks
 from lightning.pytorch.callbacks import (
@@ -91,7 +93,24 @@ def _configure_checkpoint_callbacks(self, enable_checkpointing: bool) -> None:
                     " but found `ModelCheckpoint` in callbacks list."
                 )
         elif enable_checkpointing:
-            self.trainer.callbacks.append(ModelCheckpoint())
+            if module_available("litmodels") and self.trainer._model_registry:
+                trainer_source = inspect.getmodule(self.trainer)
+                if trainer_source is None or not isinstance(trainer_source.__package__, str):
+                    raise RuntimeError("Unable to determine the source of the trainer.")
+                # this need to imported based on the actual package lightning/pytorch_lightning
+                if "pytorch_lightning" in trainer_source.__package__:
+                    from litmodels.integrations.checkpoints import PytorchLightningModelCheckpoint as LitModelCheckpoint
+                else:
+                    from litmodels.integrations.checkpoints import LightningModelCheckpoint as LitModelCheckpoint
+
+                model_checkpoint = LitModelCheckpoint(model_name=self.trainer._model_registry)
+            else:
+                rank_zero_info(
+                    "You are using the plain ModelCheckpoint callback."
+                    " Consider using LitModelCheckpoint which with seamless uploading to Model registry."
+                )
+                model_checkpoint = ModelCheckpoint()
+            self.trainer.callbacks.append(model_checkpoint)
 
     def _configure_model_summary_callback(self, enable_model_summary: bool) -> None:
         if not enable_model_summary:
diff --git a/src/lightning/pytorch/trainer/connectors/checkpoint_connector.py b/src/lightning/pytorch/trainer/connectors/checkpoint_connector.py
diff --git a/src/lightning/pytorch/trainer/trainer.py b/src/lightning/pytorch/trainer/trainer.py
diff --git a/src/lightning/pytorch/utilities/model_registry.py b/src/lightning/pytorch/utilities/model_registry.py
diff --git a/src/version.info b/src/version.info
diff --git a/tests/legacy/back-compatible-versions.txt b/tests/legacy/back-compatible-versions.txt
diff --git a/tests/tests_fabric/accelerators/test_registry.py b/tests/tests_fabric/accelerators/test_registry.py
diff --git a/tests/tests_pytorch/accelerators/test_registry.py b/tests/tests_pytorch/accelerators/test_registry.py

-Original file line number
+Diff line change
 jobs:
   check-md-links:
 -    uses: Lightning-AI/utilities/.github/workflows/[email protected].0
 +    uses: Lightning-AI/utilities/.github/workflows/[email protected].2
     with:
       config-file: ".github/markdown-links-config.json"
       base-branch: "master"