bump python 3.9+ (#20413)

* bump python 3.9+

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* --unsafe-fixes

* contextlib.AbstractContextManager

* type: ignore[misc]

* update CI

* apply fixes

* apply fixes

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Luca Antiga <[email protected]>

Author: 3 people

.actions/assistant.py

@@ -18,10 +18,11 @@
 import shutil
 import tempfile
 import urllib.request
+from collections.abc import Iterable, Iterator, Sequence
 from itertools import chain
 from os.path import dirname, isfile
 from pathlib import Path
-from typing import Any, Dict, Iterable, Iterator, List, Optional, Sequence, Tuple
+from typing import Any, Optional
 
 from packaging.requirements import Requirement
 from packaging.version import Version
@@ -127,7 +128,7 @@ def _parse_requirements(lines: Iterable[str]) -> Iterator[_RequirementWithCommen
         pip_argument = None
 
 
-def load_requirements(path_dir: str, file_name: str = "base.txt", unfreeze: str = "all") -> List[str]:
+def load_requirements(path_dir: str, file_name: str = "base.txt", unfreeze: str = "all") -> list[str]:
     """Loading requirements from a file.
 
     >>> path_req = os.path.join(_PROJECT_ROOT, "requirements")
@@ -222,7 +223,7 @@ def _load_aggregate_requirements(req_dir: str = "requirements", freeze_requireme
         fp.writelines([ln + os.linesep for ln in requires] + [os.linesep])
 
 
-def _retrieve_files(directory: str, *ext: str) -> List[str]:
+def _retrieve_files(directory: str, *ext: str) -> list[str]:
     all_files = []
     for root, _, files in os.walk(directory):
         for fname in files:
@@ -232,7 +233,7 @@ def _retrieve_files(directory: str, *ext: str) -> List[str]:
     return all_files
 
 
-def _replace_imports(lines: List[str], mapping: List[Tuple[str, str]], lightning_by: str = "") -> List[str]:
+def _replace_imports(lines: list[str], mapping: list[tuple[str, str]], lightning_by: str = "") -> list[str]:
     """Replace imports of standalone package to lightning.
 
     >>> lns = [
@@ -320,7 +321,7 @@ def copy_replace_imports(
             fo.writelines(lines)
 
 
-def create_mirror_package(source_dir: str, package_mapping: Dict[str, str]) -> None:
+def create_mirror_package(source_dir: str, package_mapping: dict[str, str]) -> None:
     """Create a mirror package with adjusted imports."""
     # replace imports and copy the code
     mapping = package_mapping.copy()

.github/workflows/_legacy-checkpoints.yml

@@ -60,7 +60,7 @@ jobs:
       - uses: actions/setup-python@v5
         with:
           # Python version here needs to be supported by all PL versions listed in back-compatible-versions.txt.
-          python-version: 3.8
+          python-version: "3.9"
 
       - name: Install PL from source
         env:

.pre-commit-config.yaml

@@ -74,7 +74,7 @@ repos:
     hooks:
       # try to fix what is possible
       - id: ruff
-        args: ["--fix"]
+        args: ["--fix", "--unsafe-fixes"]
       # perform formatting updates
       - id: ruff-format
       # validate if all is fine with preview mode

docs/source-pytorch/accelerators/tpu_faq.rst

@@ -40,9 +40,9 @@ Unsupported datatype transfer to TPUs?
 
 .. code-block::
 
-    File "/usr/local/lib/python3.8/dist-packages/torch_xla/utils/utils.py", line 205, in _for_each_instance_rewrite
+    File "/usr/local/lib/python3.9/dist-packages/torch_xla/utils/utils.py", line 205, in _for_each_instance_rewrite
         v = _for_each_instance_rewrite(result.__dict__[k], select_fn, fn, rwmap)
-    File "/usr/local/lib/python3.8/dist-packages/torch_xla/utils/utils.py", line 206, in _for_each_instance_rewrite
+    File "/usr/local/lib/python3.9/dist-packages/torch_xla/utils/utils.py", line 206, in _for_each_instance_rewrite
         result.__dict__[k] = v
     TypeError: 'mappingproxy' object does not support item assignment
 

docs/source-pytorch/advanced/post_training_quantization.rst

@@ -33,7 +33,7 @@ Installation
 Prerequisites
 =============
 
-Python version: 3.8, 3.9, 3.10
+Python version: 3.9, 3.10
 
 Install Intel® Neural Compressor
 ================================

examples/fabric/build_your_own_trainer/trainer.py

@@ -1,7 +1,7 @@
 import os
-from collections.abc import Mapping
+from collections.abc import Iterable, Mapping
 from functools import partial
-from typing import Any, Iterable, List, Literal, Optional, Tuple, Union, cast
+from typing import Any, Literal, Optional, Union, cast
 
 import lightning as L
 import torch
@@ -19,11 +19,11 @@ def __init__(
         self,
         accelerator: Union[str, Accelerator] = "auto",
         strategy: Union[str, Strategy] = "auto",
-        devices: Union[List[int], str, int] = "auto",
+        devices: Union[list[int], str, int] = "auto",
         precision: Union[str, int] = "32-true",
         plugins: Optional[Union[str, Any]] = None,
-        callbacks: Optional[Union[List[Any], Any]] = None,
-        loggers: Optional[Union[Logger, List[Logger]]] = None,
+        callbacks: Optional[Union[list[Any], Any]] = None,
+        loggers: Optional[Union[Logger, list[Logger]]] = None,
         max_epochs: Optional[int] = 1000,
         max_steps: Optional[int] = None,
         grad_accum_steps: int = 1,
@@ -465,7 +465,7 @@ def get_latest_checkpoint(checkpoint_dir: str) -> Optional[str]:
 
     def _parse_optimizers_schedulers(
         self, configure_optim_output
-    ) -> Tuple[
+    ) -> tuple[
         Optional[L.fabric.utilities.types.Optimizable],
         Optional[Mapping[str, Union[L.fabric.utilities.types.LRScheduler, bool, str, int]]],
     ]:

examples/fabric/reinforcement_learning/rl/agent.py

@@ -1,5 +1,4 @@
 import math
-from typing import Dict, Tuple
 
 import gymnasium as gym
 import torch
@@ -43,7 +42,7 @@ def __init__(self, envs: gym.vector.SyncVectorEnv, act_fun: str = "relu", ortho_
             layer_init(torch.nn.Linear(64, envs.single_action_space.n), std=0.01, ortho_init=ortho_init),
         )
 
-    def get_action(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor]:
+    def get_action(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor]:
         logits = self.actor(x)
         distribution = Categorical(logits=logits)
         if action is None:
@@ -58,12 +57,12 @@ def get_greedy_action(self, x: Tensor) -> Tensor:
     def get_value(self, x: Tensor) -> Tensor:
         return self.critic(x)
 
-    def get_action_and_value(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor, Tensor]:
+    def get_action_and_value(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor, Tensor]:
         action, log_prob, entropy = self.get_action(x, action)
         value = self.get_value(x)
         return action, log_prob, entropy, value
 
-    def forward(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor, Tensor]:
+    def forward(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor, Tensor]:
         return self.get_action_and_value(x, action)
 
     @torch.no_grad()
@@ -77,7 +76,7 @@ def estimate_returns_and_advantages(
         num_steps: int,
         gamma: float,
         gae_lambda: float,
-    ) -> Tuple[Tensor, Tensor]:
+    ) -> tuple[Tensor, Tensor]:
         next_value = self.get_value(next_obs).reshape(1, -1)
         advantages = torch.zeros_like(rewards)
         lastgaelam = 0
@@ -143,7 +142,7 @@ def __init__(
         self.avg_value_loss = MeanMetric(**torchmetrics_kwargs)
         self.avg_ent_loss = MeanMetric(**torchmetrics_kwargs)
 
-    def get_action(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor]:
+    def get_action(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor]:
         logits = self.actor(x)
         distribution = Categorical(logits=logits)
         if action is None:
@@ -158,12 +157,12 @@ def get_greedy_action(self, x: Tensor) -> Tensor:
     def get_value(self, x: Tensor) -> Tensor:
         return self.critic(x)
 
-    def get_action_and_value(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor, Tensor]:
+    def get_action_and_value(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor, Tensor]:
         action, log_prob, entropy = self.get_action(x, action)
         value = self.get_value(x)
         return action, log_prob, entropy, value
 
-    def forward(self, x: Tensor, action: Tensor = None) -> Tuple[Tensor, Tensor, Tensor, Tensor]:
+    def forward(self, x: Tensor, action: Tensor = None) -> tuple[Tensor, Tensor, Tensor, Tensor]:
         return self.get_action_and_value(x, action)
 
     @torch.no_grad()
@@ -177,7 +176,7 @@ def estimate_returns_and_advantages(
         num_steps: int,
         gamma: float,
         gae_lambda: float,
-    ) -> Tuple[Tensor, Tensor]:
+    ) -> tuple[Tensor, Tensor]:
         next_value = self.get_value(next_obs).reshape(1, -1)
         advantages = torch.zeros_like(rewards)
         lastgaelam = 0
@@ -193,7 +192,7 @@ def estimate_returns_and_advantages(
         returns = advantages + values
         return returns, advantages
 
-    def training_step(self, batch: Dict[str, Tensor]):
+    def training_step(self, batch: dict[str, Tensor]):
         # Get actions and values given the current observations
         _, newlogprob, entropy, newvalue = self(batch["obs"], batch["actions"].long())
         logratio = newlogprob - batch["logprobs"]

examples/fabric/reinforcement_learning/train_fabric.py

@@ -21,7 +21,6 @@
 import os
 import time
 from datetime import datetime
-from typing import Dict
 
 import gymnasium as gym
 import torch
@@ -38,7 +37,7 @@ def train(
     fabric: Fabric,
     agent: PPOLightningAgent,
     optimizer: torch.optim.Optimizer,
-    data: Dict[str, Tensor],
+    data: dict[str, Tensor],
     global_step: int,
     args: argparse.Namespace,
 ):

examples/fabric/reinforcement_learning/train_torch.py

@@ -22,7 +22,6 @@
 import random
 import time
 from datetime import datetime
-from typing import Dict
 
 import gymnasium as gym
 import torch
@@ -41,7 +40,7 @@
 def train(
     agent: PPOAgent,
     optimizer: torch.optim.Optimizer,
-    data: Dict[str, Tensor],
+    data: dict[str, Tensor],
     logger: SummaryWriter,
     global_step: int,
     args: argparse.Namespace,

examples/fabric/tensor_parallel/model.py

@@ -9,7 +9,7 @@
 
 
 from dataclasses import dataclass
-from typing import Optional, Tuple
+from typing import Optional
 
 import torch
 import torch.nn.functional as F
@@ -87,7 +87,7 @@ def apply_rotary_emb(
     xq: torch.Tensor,
     xk: torch.Tensor,
     freqs_cis: torch.Tensor,
-) -> Tuple[torch.Tensor, torch.Tensor]:
+) -> tuple[torch.Tensor, torch.Tensor]:
     """Apply rotary embeddings to input tensors using the given frequency tensor.
 
     This function applies rotary embeddings to the given query 'xq' and key 'xk' tensors using the provided