chore: bump PyTorch version in dependencies & CI

.azure/gpu-tests-fabric.yml

@@ -60,13 +60,13 @@ jobs:
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.10-torch2.1-cuda12.1.1"
           PACKAGE_NAME: "fabric"
         "Fabric | latest":
-          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.7-cuda12.6.3"
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.8-cuda12.6.3"
           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-torch2.7-cuda12.6.3"
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.8-cuda12.6.3"
           PACKAGE_NAME: "lightning"
     workspace:
       clean: all

.azure/gpu-tests-pytorch.yml

@@ -53,13 +53,13 @@ jobs:
           image: "pytorchlightning/pytorch_lightning:base-cuda-py3.10-torch2.1-cuda12.1.1"
           PACKAGE_NAME: "pytorch"
         "PyTorch | latest":
-          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.7-cuda12.6.3"
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.8-cuda12.6.3"
           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-torch2.7-cuda12.6.3"
+          image: "pytorchlightning/pytorch_lightning:base-cuda-py3.12-torch2.8-cuda12.6.3"
           PACKAGE_NAME: "lightning"
     pool: lit-rtx-3090
     variables:

.github/workflows/ci-tests-fabric.yml

@@ -64,13 +64,13 @@ jobs:
           - { os: "ubuntu-22.04", pkg-name: "fabric", pytorch-version: "2.1", requires: "oldest" }
           - { os: "windows-2022", pkg-name: "fabric", pytorch-version: "2.1", requires: "oldest" }
           # "fabric" installs the standalone package
-          - { os: "macOS-14", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.5" }
-          - { os: "ubuntu-22.04", 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" }
+          - { os: "macOS-14", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.7" }
+          - { os: "ubuntu-22.04", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.7" }
+          - { os: "windows-2022", pkg-name: "fabric", python-version: "3.10", pytorch-version: "2.7" }
           # 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: "2.8" }
+          - { os: "ubuntu-22.04", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.8" }
+          - { os: "windows-2022", pkg-name: "fabric", python-version: "3.12", pytorch-version: "2.8" }
     timeout-minutes: 25 # because of building grpcio on Mac
     env:
       PACKAGE_NAME: ${{ matrix.pkg-name }}

.github/workflows/ci-tests-pytorch.yml

@@ -68,13 +68,13 @@ jobs:
           - { os: "ubuntu-22.04", pkg-name: "pytorch", pytorch-version: "2.1", requires: "oldest" }
           - { os: "windows-2022", pkg-name: "pytorch", pytorch-version: "2.1", requires: "oldest" }
           # "pytorch" installs the standalone package
-          - { os: "macOS-14", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.5" }
-          - { os: "ubuntu-22.04", 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" }
+          - { os: "macOS-14", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.7" }
+          - { os: "ubuntu-22.04", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.7" }
+          - { os: "windows-2022", pkg-name: "pytorch", python-version: "3.10", pytorch-version: "2.7" }
           # 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: "2.8" }
+          - { os: "ubuntu-22.04", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.8" }
+          - { os: "windows-2022", pkg-name: "pytorch", python-version: "3.12", pytorch-version: "2.8" }
     timeout-minutes: 50
     env:
       PACKAGE_NAME: ${{ matrix.pkg-name }}

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, <0.23.0
+torchvision >=0.16.0, <0.24.0
 torchmetrics >=0.10.0, <1.9.0

requirements/pytorch/base.txt

@@ -1,7 +1,7 @@
 # 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, <=2.8.0
+torch >=2.1.0, <2.9.0
 tqdm >=4.57.0, <4.68.0
 PyYAML >5.4, <6.1.0
 fsspec[http] >=2022.5.0, <2025.8.0

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, <0.23.0
+torchvision >=0.16.0, <0.24.0
 ipython[all] <8.19.0
 torchmetrics >=0.10.0, <1.9.0

src/lightning/fabric/utilities/imports.py

@@ -36,5 +36,5 @@
 _TORCH_GREATER_EQUAL_2_4_1 = compare_version("torch", operator.ge, "2.4.1")
 _TORCH_GREATER_EQUAL_2_5 = compare_version("torch", operator.ge, "2.5.0")
 _TORCH_LESS_EQUAL_2_6 = compare_version("torch", operator.le, "2.6.0")
-
+_TORCHMETRICS_GREATER_EQUAL_1_0_0 = compare_version("torchmetrics", operator.ge, "1.0.0")
 _PYTHON_GREATER_EQUAL_3_10_0 = (sys.version_info.major, sys.version_info.minor) >= (3, 10)

src/lightning/fabric/utilities/spike.py

@@ -1,19 +1,16 @@
 import json
-import operator
 import os
 import warnings
 from typing import TYPE_CHECKING, Any, Literal, Optional, Union
 
 import torch
-from lightning_utilities.core.imports import compare_version
 
+from lightning.fabric.utilities.imports import _TORCHMETRICS_GREATER_EQUAL_1_0_0
 from lightning.fabric.utilities.types import _PATH
 
 if TYPE_CHECKING:
     from lightning.fabric.fabric import Fabric
 
-_TORCHMETRICS_GREATER_EQUAL_1_0_0 = compare_version("torchmetrics", operator.ge, "1.0.0")
-
 
 class SpikeDetection:
     """Spike Detection Callback.

src/lightning/fabric/utilities/testing/_runif.py

@@ -40,6 +40,7 @@ def _runif_reasons(
     standalone: bool = False,
     deepspeed: bool = False,
     dynamo: bool = False,
+    linux_only: bool = False,
 ) -> tuple[list[str], dict[str, bool]]:
     """Construct reasons for pytest skipif.
 
@@ -121,4 +122,7 @@ def _runif_reasons(
         if not is_dynamo_supported():
             reasons.append("torch.dynamo")
 
+    if linux_only and sys.platform != "linux":
+        reasons.append("only linux")
+
     return reasons, kwargs

src/lightning/pytorch/trainer/connectors/logger_connector/result.py

@@ -25,9 +25,9 @@
 from lightning.fabric.utilities import move_data_to_device
 from lightning.fabric.utilities.apply_func import convert_tensors_to_scalars
 from lightning.fabric.utilities.distributed import _distributed_is_initialized
+from lightning.fabric.utilities.imports import _TORCHMETRICS_GREATER_EQUAL_1_0_0
 from lightning.pytorch.utilities.data import extract_batch_size
 from lightning.pytorch.utilities.exceptions import MisconfigurationException
-from lightning.pytorch.utilities.imports import _TORCHMETRICS_GREATER_EQUAL_1_0_0
 from lightning.pytorch.utilities.memory import recursive_detach
 from lightning.pytorch.utilities.rank_zero import WarningCache, rank_zero_warn
 from lightning.pytorch.utilities.warnings import PossibleUserWarning

src/lightning/pytorch/utilities/imports.py

@@ -27,6 +27,7 @@
 _TORCHMETRICS_GREATER_EQUAL_0_9_1 = RequirementCache("torchmetrics>=0.9.1")
 _TORCHMETRICS_GREATER_EQUAL_0_11 = RequirementCache("torchmetrics>=0.11.0")  # using new API with task
 _TORCHMETRICS_GREATER_EQUAL_1_0_0 = RequirementCache("torchmetrics>=1.0.0")
+_TORCH_EQUAL_2_8 = RequirementCache("torch>=2.8.0,<2.9.0")
 
 _OMEGACONF_AVAILABLE = package_available("omegaconf")
 _TORCHVISION_AVAILABLE = RequirementCache("torchvision")

src/lightning/pytorch/utilities/testing/_runif.py

@@ -15,7 +15,7 @@
 
 from lightning_utilities.core.imports import RequirementCache
 
-from lightning.fabric.utilities.testing import _runif_reasons as fabric_run_if
+from lightning.fabric.utilities.testing import _runif_reasons as _fabric_run_if
 from lightning.pytorch.accelerators.cpu import _PSUTIL_AVAILABLE
 from lightning.pytorch.callbacks.progress.rich_progress import _RICH_AVAILABLE
 from lightning.pytorch.core.module import _ONNX_AVAILABLE, _ONNXSCRIPT_AVAILABLE
@@ -42,6 +42,7 @@ def _runif_reasons(
     psutil: bool = False,
     sklearn: bool = False,
     onnx: bool = False,
+    linux_only: bool = False,
     onnxscript: bool = False,
 ) -> tuple[list[str], dict[str, bool]]:
     """Construct reasons for pytest skipif.
@@ -69,7 +70,7 @@ def _runif_reasons(
 
     """
 
-    reasons, kwargs = fabric_run_if(
+    reasons, kwargs = _fabric_run_if(
         min_cuda_gpus=min_cuda_gpus,
         min_torch=min_torch,
         max_torch=max_torch,
@@ -81,6 +82,7 @@ def _runif_reasons(
         standalone=standalone,
         deepspeed=deepspeed,
         dynamo=dynamo,
+        linux_only=linux_only,
     )
 
     if rich and not _RICH_AVAILABLE:

tests/tests_fabric/plugins/precision/test_amp_integration.py

@@ -42,8 +42,8 @@ def forward(self, x):
 @pytest.mark.parametrize(
     ("accelerator", "precision", "expected_dtype"),
     [
-        ("cpu", "16-mixed", torch.bfloat16),
-        ("cpu", "bf16-mixed", torch.bfloat16),
+        pytest.param("cpu", "16-mixed", torch.bfloat16, marks=RunIf(skip_windows=True)),
+        pytest.param("cpu", "bf16-mixed", torch.bfloat16, marks=RunIf(skip_windows=True)),
         pytest.param("cuda", "16-mixed", torch.float16, marks=RunIf(min_cuda_gpus=2)),
         pytest.param("cuda", "bf16-mixed", torch.bfloat16, marks=RunIf(min_cuda_gpus=2, bf16_cuda=True)),
     ],

tests/tests_fabric/strategies/launchers/test_multiprocessing_integration.py

@@ -29,7 +29,8 @@ def __init__(self):
         self.register_buffer("buffer", torch.ones(3))
 
 
-@pytest.mark.parametrize("strategy", ["ddp_spawn", pytest.param("ddp_fork", marks=RunIf(skip_windows=True))])
+@RunIf(skip_windows=True)
+@pytest.mark.parametrize("strategy", ["ddp_spawn", "ddp_fork"])
 def test_memory_sharing_disabled(strategy):
     """Test that the multiprocessing launcher disables memory sharing on model parameters and buffers to avoid race
     conditions on model updates."""

tests/tests_fabric/strategies/test_ddp_integration.py

@@ -36,7 +36,7 @@
 @pytest.mark.parametrize(
     "accelerator",
     [
-        "cpu",
+        pytest.param("cpu", marks=RunIf(skip_windows=True)),
         pytest.param("cuda", marks=RunIf(min_cuda_gpus=2)),
     ],
 )

tests/tests_fabric/utilities/test_distributed.py

@@ -128,9 +128,10 @@ def test_collective_operations(devices, process):
 
 
 @pytest.mark.skipif(
-    RequirementCache("torch<2.4") and RequirementCache("numpy>=2.0"),
+    RequirementCache("numpy>=2.0"),
     reason="torch.distributed not compatible with numpy>=2.0",
 )
+@RunIf(min_torch="2.4", skip_windows=True)
 @pytest.mark.flaky(reruns=3)  # flaky with "process 0 terminated with signal SIGABRT" (GLOO)
 def test_is_shared_filesystem(tmp_path, monkeypatch):
     # In the non-distributed case, every location is interpreted as 'shared'