pytorch-lightning 2.6.0.dev0__tar.gz → 2.6.1__tar.gz

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/.actions/assistant.py

@@ -46,14 +46,14 @@ _PROJECT_ROOT = os.path.dirname(os.path.dirname(__file__))
 
 
 class _RequirementWithComment(Requirement):
-    strict_string = "# strict"
+    strict_cmd = "strict"
 
     def __init__(self, *args: Any, comment: str = "", pip_argument: Optional[str] = None, **kwargs: Any) -> None:
         super().__init__(*args, **kwargs)
         self.comment = comment
         assert pip_argument is None or pip_argument  # sanity check that it's not an empty str
         self.pip_argument = pip_argument
-        self.strict = self.strict_string in comment.lower()
+        self.strict = self.strict_cmd in comment.lower()
 
     def adjust(self, unfreeze: str) -> str:
         """Remove version restrictions unless they are strict.
@@ -62,25 +62,26 @@ class _RequirementWithComment(Requirement):
         'arrow<=1.2.2,>=1.2.0'
         >>> _RequirementWithComment("arrow<=1.2.2,>=1.2.0", comment="# strict").adjust("none")
         'arrow<=1.2.2,>=1.2.0  # strict'
-        >>> _RequirementWithComment("arrow<=1.2.2,>=1.2.0", comment="# my name").adjust("all")
-        'arrow>=1.2.0'
+        >>> _RequirementWithComment('arrow<=1.2.2,>=1.2.0; python_version >= "3.10"', comment="# my name").adjust("all")
+        'arrow>=1.2.0; python_version >= "3.10"'
         >>> _RequirementWithComment("arrow>=1.2.0, <=1.2.2", comment="# strict").adjust("all")
         'arrow<=1.2.2,>=1.2.0  # strict'
-        >>> _RequirementWithComment("arrow").adjust("all")
-        'arrow'
+        >>> _RequirementWithComment('arrow; python_version >= "3.10"').adjust("all")
+        'arrow; python_version >= "3.10"'
         >>> _RequirementWithComment("arrow>=1.2.0, <=1.2.2", comment="# cool").adjust("major")
         'arrow<2.0,>=1.2.0'
         >>> _RequirementWithComment("arrow>=1.2.0, <=1.2.2", comment="# strict").adjust("major")
         'arrow<=1.2.2,>=1.2.0  # strict'
-        >>> _RequirementWithComment("arrow>=1.2.0").adjust("major")
-        'arrow>=1.2.0'
+        >>> _RequirementWithComment('arrow>=1.2.0; python_version >= "3.10"').adjust("major")
+        'arrow>=1.2.0; python_version >= "3.10"'
         >>> _RequirementWithComment("arrow").adjust("major")
         'arrow'
 
         """
         out = str(self)
         if self.strict:
-            return f"{out}  {self.strict_string}"
+            return f"{out}  # {self.strict_cmd}"
+
         specs = [(spec.operator, spec.version) for spec in self.specifier]
         if unfreeze == "major":
             for operator, version in specs:

{pytorch_lightning-2.6.0.dev0/src/pytorch_lightning.egg-info → pytorch_lightning-2.6.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytorch-lightning
-Version: 2.6.0.dev0
+Version: 2.6.1
 Summary: PyTorch Lightning is the lightweight PyTorch wrapper for ML researchers. Scale your models. Write less boilerplate.
 Home-page: https://github.com/Lightning-AI/lightning
 Download-URL: https://github.com/Lightning-AI/lightning
@@ -21,11 +21,11 @@ Classifier: Topic :: Scientific/Engineering :: Information Analysis
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
-Requires-Python: >=3.9
+Classifier: Programming Language :: Python :: 3.13
+Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: torch>=2.1.0
@@ -33,7 +33,7 @@ Requires-Dist: tqdm>=4.57.0
 Requires-Dist: PyYAML>5.4
 Requires-Dist: fsspec[http]>=2022.5.0
 Requires-Dist: torchmetrics>0.7.0
-Requires-Dist: packaging>=20.0
+Requires-Dist: packaging>=23.0
 Requires-Dist: typing-extensions>4.5.0
 Requires-Dist: lightning-utilities>=0.10.0
 Provides-Extra: extra
@@ -46,32 +46,34 @@ Requires-Dist: tensorboardX>=2.2; extra == "extra"
 Requires-Dist: bitsandbytes>=0.45.2; platform_system != "Darwin" and extra == "extra"
 Provides-Extra: strategies
 Requires-Dist: deepspeed<0.17.0,>=0.15.0; (platform_system != "Windows" and platform_system != "Darwin") and extra == "strategies"
-Provides-Extra: examples
-Requires-Dist: requests<2.33.0; extra == "examples"
-Requires-Dist: torchvision>=0.16.0; extra == "examples"
-Requires-Dist: ipython[all]>=8.0.0; extra == "examples"
-Requires-Dist: torchmetrics>=0.10.0; extra == "examples"
 Provides-Extra: test
-Requires-Dist: coverage==7.10.7; extra == "test"
-Requires-Dist: pytest==8.4.2; extra == "test"
-Requires-Dist: pytest-cov==6.3.0; extra == "test"
+Requires-Dist: coverage==7.13.1; python_version >= "3.10" and extra == "test"
+Requires-Dist: coverage==7.10.7; python_version < "3.10" and extra == "test"
+Requires-Dist: pytest==9.0.2; extra == "test"
+Requires-Dist: pytest-cov==7.0.0; extra == "test"
 Requires-Dist: pytest-timeout==2.4.0; extra == "test"
 Requires-Dist: pytest-rerunfailures==16.0.1; python_version < "3.10" and extra == "test"
 Requires-Dist: pytest-rerunfailures==16.1; python_version >= "3.10" and extra == "test"
 Requires-Dist: pytest-random-order==1.2.0; extra == "test"
 Requires-Dist: cloudpickle>=1.3; extra == "test"
 Requires-Dist: scikit-learn>0.22.1; extra == "test"
-Requires-Dist: numpy>1.20.0; extra == "test"
+Requires-Dist: numpy>1.21.0; python_version < "3.12" and extra == "test"
+Requires-Dist: numpy>2.1.0; python_version >= "3.12" and extra == "test"
 Requires-Dist: onnx>1.12.0; extra == "test"
 Requires-Dist: onnxruntime>=1.12.0; extra == "test"
 Requires-Dist: onnxscript>=0.1.0; extra == "test"
-Requires-Dist: psutil<7.1.2; extra == "test"
+Requires-Dist: psutil<7.3.0; extra == "test"
 Requires-Dist: pandas>2.0; extra == "test"
 Requires-Dist: fastapi; extra == "test"
 Requires-Dist: uvicorn; extra == "test"
 Requires-Dist: tensorboard>=2.11; extra == "test"
-Requires-Dist: torch-tensorrt; (platform_system == "Linux" and python_version >= "3.12") and extra == "test"
+Requires-Dist: torch-tensorrt; (platform_system != "Darwin" and python_version >= "3.12") and extra == "test"
 Requires-Dist: huggingface-hub; extra == "test"
+Provides-Extra: examples
+Requires-Dist: requests<2.33.0; extra == "examples"
+Requires-Dist: torchvision>=0.16.0; extra == "examples"
+Requires-Dist: ipython[all]>=8.0.0; extra == "examples"
+Requires-Dist: torchmetrics>=0.10.0; extra == "examples"
 Provides-Extra: deepspeed
 Requires-Dist: deepspeed<0.17.0,>=0.15.0; (platform_system != "Windows" and platform_system != "Darwin") and extra == "deepspeed"
 Provides-Extra: all
@@ -100,25 +102,27 @@ Requires-Dist: requests<2.33.0; extra == "dev"
 Requires-Dist: torchvision>=0.16.0; extra == "dev"
 Requires-Dist: ipython[all]>=8.0.0; extra == "dev"
 Requires-Dist: torchmetrics>=0.10.0; extra == "dev"
-Requires-Dist: coverage==7.10.7; extra == "dev"
-Requires-Dist: pytest==8.4.2; extra == "dev"
-Requires-Dist: pytest-cov==6.3.0; extra == "dev"
+Requires-Dist: coverage==7.13.1; python_version >= "3.10" and extra == "dev"
+Requires-Dist: coverage==7.10.7; python_version < "3.10" and extra == "dev"
+Requires-Dist: pytest==9.0.2; extra == "dev"
+Requires-Dist: pytest-cov==7.0.0; extra == "dev"
 Requires-Dist: pytest-timeout==2.4.0; extra == "dev"
 Requires-Dist: pytest-rerunfailures==16.0.1; python_version < "3.10" and extra == "dev"
 Requires-Dist: pytest-rerunfailures==16.1; python_version >= "3.10" and extra == "dev"
 Requires-Dist: pytest-random-order==1.2.0; extra == "dev"
 Requires-Dist: cloudpickle>=1.3; extra == "dev"
 Requires-Dist: scikit-learn>0.22.1; extra == "dev"
-Requires-Dist: numpy>1.20.0; extra == "dev"
+Requires-Dist: numpy>1.21.0; python_version < "3.12" and extra == "dev"
+Requires-Dist: numpy>2.1.0; python_version >= "3.12" and extra == "dev"
 Requires-Dist: onnx>1.12.0; extra == "dev"
 Requires-Dist: onnxruntime>=1.12.0; extra == "dev"
 Requires-Dist: onnxscript>=0.1.0; extra == "dev"
-Requires-Dist: psutil<7.1.2; extra == "dev"
+Requires-Dist: psutil<7.3.0; extra == "dev"
 Requires-Dist: pandas>2.0; extra == "dev"
 Requires-Dist: fastapi; extra == "dev"
 Requires-Dist: uvicorn; extra == "dev"
 Requires-Dist: tensorboard>=2.11; extra == "dev"
-Requires-Dist: torch-tensorrt; (platform_system == "Linux" and python_version >= "3.12") and extra == "dev"
+Requires-Dist: torch-tensorrt; (platform_system != "Darwin" and python_version >= "3.12") and extra == "dev"
 Requires-Dist: huggingface-hub; extra == "dev"
 Dynamic: author
 Dynamic: author-email
@@ -164,7 +168,7 @@ ______________________________________________________________________
 [![DockerHub](https://img.shields.io/docker/pulls/pytorchlightning/pytorch_lightning.svg)](https://hub.docker.com/r/pytorchlightning/pytorch_lightning)
 [![codecov](https://codecov.io/gh/Lightning-AI/pytorch-lightning/graph/badge.svg?token=SmzX8mnKlA)](https://codecov.io/gh/Lightning-AI/pytorch-lightning)
 
-[![ReadTheDocs](https://readthedocs.org/projects/pytorch-lightning/badge/?version=2.6.0dev0)](https://lightning.ai/docs/pytorch/stable/)[![Discord](https://img.shields.io/discord/1077906959069626439?style=plastic)](https://discord.gg/VptPCZkGNa)
+[![ReadTheDocs](https://readthedocs.org/projects/pytorch-lightning/badge/?version=2.6.1)](https://lightning.ai/docs/pytorch/stable/)[![Discord](https://img.shields.io/discord/1077906959069626439?style=plastic)](https://discord.gg/VptPCZkGNa)
 [![license](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://github.com/Lightning-AI/pytorch-lightning/blob/master/LICENSE)
 
 <!--

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/README.md

@@ -7,7 +7,7 @@
 
 **The deep learning framework to pretrain and finetune AI models.**
 
-**Deploying models?** Check out [LitServe](https://github.com/Lightning-AI/litserve?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme), the PyTorch Lightning for inference engines
+**Serving models?** Use [LitServe](https://github.com/Lightning-AI/litserve?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme) to build custom inference servers in pure Python.
 
 ______________________________________________________________________
 
@@ -55,22 +55,17 @@ ______________________________________________________________________
 
  
 
-# Looking for GPUs?
-Over 340,000 developers use [Lightning Cloud](https://lightning.ai/?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme) - purpose-built for PyTorch and PyTorch Lightning. 
-- [GPUs](https://lightning.ai/pricing?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme) from $0.19.   
-- [Clusters](https://lightning.ai/clusters?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme): frontier-grade training/inference clusters.   
-- [AI Studio (vibe train)](https://lightning.ai/studios?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme): workspaces where AI helps you debug, tune and vibe train.
-- [AI Studio (vibe deploy)](https://lightning.ai/studios?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme): workspaces where AI helps you optimize, and deploy models.     
-- [Notebooks](https://lightning.ai/notebooks?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme): Persistent GPU workspaces where AI helps you code and analyze.
-- [Inference](https://lightning.ai/deploy?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme): Deploy models as inference APIs.   
-
 <a id="why-pytorch-lightning"></a>
 # Why PyTorch Lightning?   
-
-Training models in plain PyTorch is tedious and error-prone - you have to manually handle things like backprop, mixed precision, multi-GPU, and distributed training, often rewriting code for every new project. PyTorch Lightning organizes PyTorch code to automate those complexities so you can focus on your model and data, while keeping full control and scaling from CPU to multi-node without changing your core code. But if you want control of those things, you can still opt into [expert-level control](#lightning-fabric-expert-control).   
+Training models in plain PyTorch requires writing and maintaining a lot of repetitive engineering code. Handling backpropagation, mixed precision, multi-GPU, and distributed training is error-prone and often reimplemented for every project. PyTorch Lightning organizes PyTorch code to automate this infrastructure while keeping full control over your model logic. You write the science. Lightning handles the engineering, and scales from CPU to multi-node GPUs without changing your core code. PyTorch experts can still opt into [expert-level control](#lightning-fabric-expert-control).   
 
 Fun analogy: If PyTorch is Javascript, PyTorch Lightning is ReactJS or NextJS.
 
+# Looking for GPUs?
+[Lightning Cloud](https://lightning.ai/?utm_source=ptl_readme&utm_medium=referral&utm_campaign=ptl_readme) is the easiest way to run PyTorch Lightning without managing infrastructure. Start training with one command and get GPUs, autoscaling, monitoring, and a free tier. No cloud setup required.
+
+You can also run PyTorch Lightning on your own hardware or cloud.
+
 # Lightning has 2 core packages
 
 [PyTorch Lightning: Train and deploy PyTorch at scale](#why-pytorch-lightning).
@@ -282,6 +277,9 @@ trainer = Trainer(precision=16)
 ```python
 from lightning import loggers
 
+# litlogger
+trainer = Trainer(logger=LitLogger())
+
 # tensorboard
 trainer = Trainer(logger=TensorBoardLogger("logs/"))
 

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/pyproject.toml

@@ -92,9 +92,6 @@ ignore = [
 "src/**" = [
     "S101",  # todo: Use of `assert` detected
     "S113",  # todo: Probable use of requests call without timeout
-    "S301",  # todo: `pickle` and modules that wrap it can be unsafe when used to deserialize untrusted data, possible security issue
-    "S324",  # todo: Probable use of insecure hash functions in `hashlib`
-    "S403",  # todo: `pickle`, `cPickle`, `dill`, and `shelve` modules are possibly insecure
     "S404",  # todo: `subprocess` module is possibly insecure
     "S602",  # todo: `subprocess` call with `shell=True` identified, security issue
     "S603",  # todo: `subprocess` call: check for execution of untrusted input
@@ -106,18 +103,14 @@ ignore = [
 "tests/**" = [
     "S101",  # Use of `assert` detected
     "S301",  # `pickle` and modules that wrap it can be unsafe when used to deserialize untrusted data, possible security issue
-    "S113",  # todo: Probable use of requests call without timeout
     "S311",  # todo: Standard pseudo-random generators are not suitable for cryptographic purposes
     "S108",  # todo: Probable insecure usage of temporary file or directory: "/tmp/sys-customizations-sync"
-    "S202",  # Uses of `tarfile.extractall()`
     "S403",  # `pickle`, `cPickle`, `dill`, and `shelve` modules are possibly insecure
     "S404",  # `subprocess` module is possibly insecure
     "S602",  # todo: `subprocess` call with `shell=True` identified, security issue
     "S603",  # todo: `subprocess` call: check for execution of untrusted input
     "S605",  # todo: Starting a process with a shell: seems safe, but may be changed in the future; consider rewriting without `shell`
     "S607",  # todo: Starting a process with a partial executable path
-    "RET504",  # todo:Unnecessary variable assignment before `return` statement
-    "PT004",  # todo: Fixture `tmpdir_unittest_fixture` does not return anything, add leading underscore
     "PT012",  # todo: `pytest.raises()` block should contain a single simple statement
     "PT019",  # todo: Fixture `_` without value is injected as parameter, use `@pytest.mark.usefixtures` instead
 ]
@@ -179,6 +172,10 @@ filterwarnings = [
     # "error::DeprecationWarning",
     "error::FutureWarning",
     "ignore::FutureWarning:onnxscript", # Temporary ignore until onnxscript is updated
+    "ignore:You are using `torch.load` with `weights_only=False`.*:FutureWarning",
+    "ignore::FutureWarning:torch.distributed.algorithms.ddp_comm_hooks",
+    # PyTorch pytree LeafSpec deprecation triggered during doctest
+    "ignore:`isinstance\\(treespec, LeafSpec\\)` is deprecated.*:FutureWarning",
 ]
 xfail_strict = true
 junit_duration_report = "call"

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/fabric/base.txt

@@ -1,8 +1,8 @@
 # NOTE: the upper bound for the package version is only set for CI stability, and it is dropped while installing this package
 #  in case you want to preserve/enforce restrictions on the latest compatible version, add "strict" as an in-line comment
 
-torch >=2.1.0, <2.9.0
-fsspec[http] >=2022.5.0, <2025.11.0
-packaging >=20.0, <=25.0
+torch >=2.1.0, <2.10.0
+fsspec[http] >=2022.5.0, <2026.2.0
+packaging >=23.0, <=25.0
 typing-extensions >4.5.0, <4.16.0
 lightning-utilities >=0.10.0, <0.16.0

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/fabric/examples.txt

@@ -1,5 +1,5 @@
 # NOTE: the upper bound for the package version is only set for CI stability, and it is dropped while installing this package
 #  in case you want to preserve/enforce restrictions on the latest compatible version, add "strict" as an in-line comment
 
-torchvision >=0.16.0, <0.25.0
+torchvision >=0.16.0, <0.26.0
 torchmetrics >=0.10.0, <1.9.0

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/fabric/test.txt

@@ -1,12 +1,14 @@
-coverage ==7.10.7
-numpy >=1.21.0, <1.27.0
-pytest ==8.4.2
-pytest-cov ==6.3.0
+coverage ==7.13.1; python_version >= "3.10"
+coverage ==7.10.7; python_version < "3.10"
+numpy >1.21.0, <1.27.0; python_version < "3.12"
+numpy >2.1.0, <2.4.2; python_version >= "3.12"
+pytest ==9.0.2
+pytest-cov ==7.0.0
 pytest-timeout ==2.4.0
 pytest-rerunfailures ==16.0.1; python_version < "3.10"
 pytest-rerunfailures ==16.1; python_version >= "3.10"
 pytest-random-order ==1.2.0
 click ==8.1.8; python_version < "3.11"
-click ==8.3.0; python_version > "3.10"
+click ==8.3.1; python_version > "3.10"
 tensorboardX >=2.6, <2.7.0  # todo: relax it back to `>=2.2` after fixing tests
 huggingface-hub

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/pytorch/base.txt

@@ -1,11 +1,11 @@
 # NOTE: the upper bound for the package version is only set for CI stability, and it is dropped while installing this package
 #  in case you want to preserve/enforce restrictions on the latest compatible version, add "strict" as an in-line comment
 
-torch >=2.1.0, <2.9.0
+torch >=2.1.0, <2.10.0
 tqdm >=4.57.0, <4.68.0
 PyYAML >5.4, <6.1.0
-fsspec[http] >=2022.5.0, <2025.11.0
+fsspec[http] >=2022.5.0, <2026.2.0
 torchmetrics >0.7.0, <1.9.0
-packaging >=20.0, <=25.0
+packaging >=23.0, <=25.0
 typing-extensions >4.5.0, <4.16.0
 lightning-utilities >=0.10.0, <0.16.0

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/pytorch/docs.txt

@@ -1,9 +1,9 @@
 -r ../docs.txt
 
 nbformat  # used for generate empty notebook
-ipython[notebook] <9.7.0
+ipython[notebook] <9.10.0
 setuptools<81.0  # workaround for `error in ipython setup command: use_2to3 is invalid.`
 
-onnxscript >= 0.2.2, < 0.5.0
+onnxscript >= 0.2.2, < 0.6.0
 
 #-r ../../_notebooks/.actions/requires.txt

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/pytorch/examples.txt

@@ -2,6 +2,6 @@
 #  in case you want to preserve/enforce restrictions on the latest compatible version, add "strict" as an in-line comment
 
 requests <2.33.0
-torchvision >=0.16.0, <0.25.0
+torchvision >=0.16.0, <0.26.0
 ipython[all] >=8.0.0, <10.0.0
 torchmetrics >=0.10.0, <1.9.0

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/pytorch/extra.txt

@@ -5,7 +5,7 @@
 matplotlib>3.1, <3.11.0
 omegaconf >=2.2.3, <2.4.0
 hydra-core >=1.2.0, <1.4.0
-jsonargparse[signatures,jsonnet] >=4.39.0, <4.43.0
-rich >=12.3.0, <14.2.0
+jsonargparse[signatures,jsonnet] >=4.39.0, <4.46.0
+rich >=12.3.0, <14.3.0
 tensorboardX >=2.2, <2.7.0  # min version is set by torch.onnx missing attribute
 bitsandbytes >=0.45.2,<0.47.0; platform_system != "Darwin"

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/requirements/pytorch/test.txt

@@ -1,6 +1,7 @@
-coverage ==7.10.7
-pytest ==8.4.2
-pytest-cov ==6.3.0
+coverage ==7.13.1; python_version >= "3.10"
+coverage ==7.10.7; python_version < "3.10"
+pytest ==9.0.2
+pytest-cov ==7.0.0
 pytest-timeout ==2.4.0
 pytest-rerunfailures ==16.0.1; python_version < "3.10"
 pytest-rerunfailures ==16.1; python_version >= "3.10"
@@ -8,17 +9,18 @@ pytest-random-order ==1.2.0
 
 # needed in tests
 cloudpickle >=1.3, <3.2.0
-scikit-learn >0.22.1, <1.8.0
-numpy >1.20.0, <1.27.0
-onnx >1.12.0, <1.20.0
+scikit-learn >0.22.1, <1.9.0
+numpy >1.21.0, <1.27.0; python_version < "3.12"
+numpy >2.1.0, <2.4.2; python_version >= "3.12"
+onnx >1.12.0, <1.21.0
 onnxruntime >=1.12.0, <1.24.0
-onnxscript >= 0.1.0, < 0.5.0
-psutil <7.1.2 # for `DeviceStatsMonitor`
+onnxscript >= 0.1.0, < 0.6.0
+psutil <7.3.0 # for `DeviceStatsMonitor`
 pandas >2.0, <2.4.0  # needed in benchmarks
 fastapi  # for `ServableModuleValidator`  # not setting version as re-defined in App
 uvicorn  # for `ServableModuleValidator`  # not setting version as re-defined in App
 
 tensorboard >=2.11, <2.21.0  # for `TensorBoardLogger`
 
-torch-tensorrt; platform_system == "Linux" and python_version >= "3.12"
+torch-tensorrt; platform_system != "Darwin" and python_version >= "3.12"
 huggingface-hub

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/src/lightning_fabric/CHANGELOG.md

@@ -22,12 +22,56 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 -
 
 
+### Fixed
+
+-
+
+---
+
+## [2.6.1] - 2026-01-30
+
+### Added
+
+- Exposed `weights_only` argument for loading checkpoints in `Fabric.load()` and `Fabric.load_raw()` ([#21470](https://github.com/Lightning-AI/pytorch-lightning/pull/21470))
+
+### Changed
+
+-
+
+### Fixed
+
+- Fixed `DistributedSamplerWrapper` not forwarding `set_epoch` to the underlying sampler ([#21454](https://github.com/Lightning-AI/pytorch-lightning/pull/21454))
+
+- Fixed DDP notebook CUDA fork check to allow passive initialization when CUDA is not actively used ([#21402](https://github.com/Lightning-AI/pytorch-lightning/pull/21402))
+
+### Removed
+
+-
+
+---
+
+
+## [2.6.0] - 2025-11-28
+
+### Added
+
+- Added kwargs-filtering for `Fabric.call` to support different callback method signatures ([#21258](https://github.com/Lightning-AI/pytorch-lightning/pull/21258))
+
+
+### Changed
+
+- Expose `weights_only` argument for `Trainer.{fit,validate,test,predict}` and let `torch` handle default value ([#21072](https://github.com/Lightning-AI/pytorch-lightning/pull/21072))
 - Set `_DeviceDtypeModuleMixin._device` from torch's default device function ([#21164](https://github.com/Lightning-AI/pytorch-lightning/pull/21164))
 
 
 ### Fixed
 
-- Fixed `EADDRINUSE` errors in distributed tests with port manager and retry logic ([#21309](https://github.com/Lightning-AI/pytorch-lightning/pull/21309))
+- Fixed issue in detecting MPIEnvironment with partial mpi4py installation ([#21353](https://github.com/Lightning-AI/pytorch-lightning/pull/21353))
+
+- Learning rate scheduler is stepped at the end of epoch when `on_train_batch_start` returns -1 ([#21296](https://github.com/Lightning-AI/pytorch-lightning/issues/21296)).
+
+
+- Fixed FSDP mixed precision semantics and added user warning ([#21361](https://github.com/Lightning-AI/pytorch-lightning/pull/21361))
 
 
 ---

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/src/lightning_fabric/__setup__.py

@@ -73,7 +73,7 @@ def _setup_args() -> dict[str, Any]:
         "include_package_data": True,
         "zip_safe": False,
         "keywords": ["deep learning", "pytorch", "AI"],
-        "python_requires": ">=3.9",
+        "python_requires": ">=3.10",
         "setup_requires": ["wheel"],
         "install_requires": assistant.load_requirements(
             _PATH_REQUIREMENTS, unfreeze="none" if _FREEZE_REQUIREMENTS else "all"
@@ -105,9 +105,9 @@ def _setup_args() -> dict[str, Any]:
             # Specify the Python versions you support here. In particular, ensure
             # that you indicate whether you support Python 2, Python 3 or both.
             "Programming Language :: Python :: 3",
-            "Programming Language :: Python :: 3.9",
             "Programming Language :: Python :: 3.10",
             "Programming Language :: Python :: 3.11",
             "Programming Language :: Python :: 3.12",
+            "Programming Language :: Python :: 3.13",
         ],
     }

{pytorch_lightning-2.6.0.dev0 → pytorch_lightning-2.6.1}/src/lightning_fabric/fabric.py

@@ -871,8 +871,10 @@ class Fabric:
         path: Union[str, Path],
         state: Optional[dict[str, Union[nn.Module, Optimizer, Any]]] = None,
         strict: bool = True,
+        *,
+        weights_only: Optional[bool] = None,
     ) -> dict[str, Any]:
-        """Load a checkpoint from a file and restore the state of objects (modules, optimizers, etc.)
+        """Load a checkpoint from a file and restore the state of objects (modules, optimizers, etc.).
 
         How and which processes load gets determined by the `strategy`.
         This method must be called on all processes!
@@ -881,7 +883,12 @@ class Fabric:
             path: A path to where the file is located.
             state: A dictionary of objects whose state will be restored in-place from the checkpoint path.
                 If no state is given, then the checkpoint will be returned in full.
-            strict: Whether to enforce that the keys in `state` match the keys in the checkpoint.
+            strict: Whether to enforce that the keys in ``state`` match the keys in the checkpoint.
+            weights_only: Defaults to ``None``. If ``True``, restricts loading to ``state_dicts`` of plain
+                ``torch.Tensor`` and other primitive types. If loading a checkpoint from a trusted source that contains
+                an ``nn.Module``, use ``weights_only=False``. If loading checkpoint from an untrusted source, we
+                recommend using ``weights_only=True``. For more information, please refer to the
+                `PyTorch Developer Notes on Serialization Semantics <https://docs.pytorch.org/docs/main/notes/serialization.html#id3>`_.
 
         Returns:
             The remaining items that were not restored into the given state dictionary. If no state dictionary is
@@ -899,7 +906,12 @@ class Fabric:
 
         """
         unwrapped_state = _unwrap_objects(state)
-        remainder = self._strategy.load_checkpoint(path=path, state=unwrapped_state, strict=strict)
+        remainder = self._strategy.load_checkpoint(
+            path=path,
+            state=unwrapped_state,
+            strict=strict,
+            weights_only=weights_only,
+        )
         self.barrier()
         if state is not None:
             # We need to unwrap objects (see above) but this creates a new dictionary. In-place updates
@@ -911,7 +923,14 @@ class Fabric:
                 state[k] = unwrapped_state[k]
         return remainder
 
-    def load_raw(self, path: Union[str, Path], obj: Union[nn.Module, Optimizer], strict: bool = True) -> None:
+    def load_raw(
+        self,
+        path: Union[str, Path],
+        obj: Union[nn.Module, Optimizer],
+        strict: bool = True,
+        *,
+        weights_only: Optional[bool] = None,
+    ) -> None:
         """Load the state of a module or optimizer from a single state-dict file.
 
         Use this for loading a raw PyTorch model checkpoint created without Fabric.
@@ -923,10 +942,15 @@ class Fabric:
             obj: A :class:`~torch.nn.Module` or :class:`~torch.optim.Optimizer` instance.
             strict: Whether to enforce that the keys in the module's state-dict match the keys in the checkpoint.
                 Does not apply to optimizers.
+            weights_only: Defaults to ``None``. If ``True``, restricts loading to ``state_dicts`` of plain
+                ``torch.Tensor`` and other primitive types. If loading a checkpoint from a trusted source that contains
+                an ``nn.Module``, use ``weights_only=False``. If loading checkpoint from an untrusted source, we
+                recommend using ``weights_only=True``. For more information, please refer to the
+                `PyTorch Developer Notes on Serialization Semantics <https://docs.pytorch.org/docs/main/notes/serialization.html#id3>`_.
 
         """
         obj = _unwrap_objects(obj)
-        self._strategy.load_checkpoint(path=path, state=obj, strict=strict)
+        self._strategy.load_checkpoint(path=path, state=obj, strict=strict, weights_only=weights_only)
 
     def launch(self, function: Callable[["Fabric"], Any] = _do_nothing, *args: Any, **kwargs: Any) -> Any:
         """Launch and initialize all the processes needed for distributed execution.
@@ -985,6 +1009,34 @@ class Fabric:
             )
         return self._wrap_and_launch(function, self, *args, **kwargs)
 
+    def _filter_kwargs_for_callback(self, method: Callable, kwargs: dict[str, Any]) -> dict[str, Any]:
+        """Filter keyword arguments to only include those that match the callback method's signature.
+
+        Args:
+            method: The callback method to inspect
+            kwargs: The keyword arguments to filter
+
+        Returns:
+            A filtered dictionary of keyword arguments that match the method's signature
+
+        """
+        try:
+            sig = inspect.signature(method)
+        except (ValueError, TypeError):
+            # If we can't inspect the signature, pass all kwargs to maintain backward compatibility
+            return kwargs
+
+        filtered_kwargs = {}
+        for name, param in sig.parameters.items():
+            # If the method accepts **kwargs, pass all original kwargs directly
+            if param.kind == inspect.Parameter.VAR_KEYWORD:
+                return kwargs
+            # If the parameter exists in the incoming kwargs, add it to filtered_kwargs
+            if name in kwargs:
+                filtered_kwargs[name] = kwargs[name]
+
+        return filtered_kwargs
+
     def call(self, hook_name: str, *args: Any, **kwargs: Any) -> None:
         r"""Trigger the callback methods with the given name and arguments.
 
@@ -994,7 +1046,9 @@ class Fabric:
         Args:
             hook_name: The name of the callback method.
             *args: Optional positional arguments that get passed down to the callback method.
-            **kwargs: Optional keyword arguments that get passed down to the callback method.
+            **kwargs: Optional keyword arguments that get passed down to the callback method. Keyword arguments
+                that are not present in the callback's signature will be filtered out automatically, allowing
+                callbacks to have different signatures for the same hook.
 
         Example::
 
@@ -1016,13 +1070,8 @@ class Fabric:
                 )
                 continue
 
-            method(*args, **kwargs)
-
-            # TODO(fabric): handle the following signatures
-            # method(self, fabric|trainer, x, y=1)
-            # method(self, fabric|trainer, *args, x, y=1)
-            # method(self, *args, y=1)
-            # method(self, *args, **kwargs)
+            filtered_kwargs = self._filter_kwargs_for_callback(method, kwargs)
+            method(*args, **filtered_kwargs)
 
     def log(self, name: str, value: Any, step: Optional[int] = None) -> None:
         """Log a scalar to all loggers that were added to Fabric.