Add device_name classmethod in Accelerator.

[GdoongMathew]

What does this PR do?

Fixes #17355

• 1 GPU

GPU available: NVIDIA GeForce RTX 3050 4GB Laptop GPU, using: 1 devices.
TPU available: False, using: 0 TPU cores
HPU available: False, using: 0 HPUs

• 2 GPUs (Kaggle)

>>> t = Trainer(devices=2)
INFO: GPU available: Tesla T4, using: 2 devices.
INFO: TPU available: False, using: 0 TPU cores
INFO: HPU available: False, using: 0 HPUs

• No cuda GPU available

GPU available: False, using: 0 devices.
TPU available: False, using: 0 TPU cores
HPU available: False, using: 0 HPUs

• MPS

GPU available: Apple M1 Pro, using: 1 devices.
TPU available: False, using: 0 TPU cores
HPU available: False, using: 0 HPUs

• XLA (Kaggle)

GPU available: False, using: 0 devices.
TPU available: v3-8, using: 8 TPU cores
HPU available: False, using: 0 HPUs

Before submitting

• Was this discussed/agreed via a GitHub issue? (not for typos and docs)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure your PR does only one thing, instead of bundling different changes together?
• Did you make sure to update the documentation with your changes? (if necessary)
• Did you write any new necessary tests? (not for typos and docs)
• Did you verify new and existing tests pass locally with your changes?
• Did you list all the breaking changes introduced by this pull request?
• Did you update the CHANGELOG? (not for typos, docs, test updates, or minor internal changes/refactors)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:

Reviewer checklist

• Is this pull request ready for review? (if not, please submit in draft mode)
• Check that all items from Before submitting are resolved
• Make sure the title is self-explanatory and the description concisely explains the PR
• Add labels and milestones (and optionally projects) to the PR so it can be classified

---

📚 Documentation preview 📚: https://pytorch-lightning--21112.org.readthedocs.build/en/21112/

[GdoongMathew]

Results in Kaggle:

• CPU: https://www.kaggle.com/code/gdoong/lightning-cpu?scriptVersionId=258513715&cellId=2
• GPU x 2: https://www.kaggle.com/code/gdoong/lightning-gpu?scriptVersionId=258514966&cellId=2
• TPU: https://www.kaggle.com/code/gdoong/lightning-tpu?scriptVersionId=258515194&cellId=2

[justusschock]

How would you handle heterogeneous gpu types? E.g. say I have one 3090 and one 4090 in my workstation. This doesn't handle it at all.

[GdoongMathew]

How would you handle heterogeneous gpu types? E.g. say I have one 3090 and one 4090 in my workstation. This doesn't handle it at all.

Hi @justusschock , based on the change, the output should look something like

GPU available: RTX 3090, RTX 4090, using: 2 devices.

that being said, if the setup is more complex - for example 2 x 3090 and 1 x 4090 - the current output might not fully reflect that.

[GdoongMathew]

Hi @Borda , a friendly ping for a review ~

[Borda]

Hi @Borda , a friendly ping for a review ~

unfortunatly, I am not in a position to approve a feature type PR, lest ping @ethanwharris

[codecov]

Codecov Report

❌ Patch coverage is 74.00000% with 13 lines in your changes missing coverage. Please review.
✅ Project coverage is 79%. Comparing base (7d2de87) to head (ea2767d).
⚠️ Report is 46 commits behind head on master.
✅ All tests successful. No failed tests found.

❗ There is a different number of reports uploaded between BASE (7d2de87) and HEAD (ea2767d). Click for more details.

HEAD has 1312 uploads less than BASE

Flag	BASE (7d2de87)	HEAD (ea2767d)
cpu	328	33
python	30	3
lightning_fabric	88	0
pytest	163	0
python3.10	30	3
lightning	150	15
python3.13	29	3
python3.12	89	9
python3.11	60	6
python3.12.7	90	9
pytorch2.1	30	6
pytest-full	165	33
pytorch2.3	15	3
pytorch2.5.1	15	3
pytorch2.2.2	15	3
pytorch2.8	30	6
pytorch_lightning	90	18
pytorch2.9	15	3
pytorch2.7	15	3
pytorch2.6	15	3
pytorch2.4.1	15	3

Additional details and impacted files

@@            Coverage Diff            @@
##           master   #21112     +/-   ##
=========================================
- Coverage      87%      79%     -8%     
=========================================
  Files         270      267      -3     
  Lines       24071    24055     -16     
=========================================
- Hits        20867    18986   -1881     
- Misses       3204     5069   +1865     

[codecov-commenter]

❌ 1 Tests Failed:

Tests completed	Failed	Passed	Skipped
3339	1	3338	531

View the full list of 1 ❄️ flaky test(s)

tests/tests_pytorch/callbacks/test_stochastic_weight_avg.py::test_swa_callback_ddp_cpu

Flake rate in main: 8.70% (Passed 21 times, Failed 2 times)

Stack Traces | 6.16s run time

tmp_path = PosixPath('.../pytest-of-runner/pytest-0/test_swa_callback_ddp_cpu0')

    @RunIf(skip_windows=True)
    def test_swa_callback_ddp_cpu(tmp_path):
>       train_with_swa(tmp_path, strategy="ddp_spawn", accelerator="cpu", devices=2)

callbacks/test_stochastic_weight_avg.py:197: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
callbacks/test_stochastic_weight_avg.py:179: in train_with_swa
    trainer.fit(model, weights_only=weights_only)
../../.venv/lib/python3.12.../pytorch_lightning/trainer/trainer.py:590: in fit
    call._call_and_handle_interrupt(
../../.venv/lib/python3.12.../pytorch_lightning/trainer/call.py:48: in _call_and_handle_interrupt
    return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
../../.venv/lib/python3.12.../strategies/launchers/multiprocessing.py:144: in launch
    while not process_context.join():
              ^^^^^^^^^^^^^^^^^^^^^^
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <torch.multiprocessing.spawn.ProcessContext object at 0x7ff7a2323b90>
timeout = None, grace_period = None

    def join(self, timeout: float | None = None, grace_period: float | None = None):
        r"""Join one or more processes within spawn context.
    
        Attempt to join one or more processes in this spawn context.
        If one of them exited with a non-zero exit status, this function
        kills the remaining processes (optionally with a grace period)
        and raises an exception with the cause of the first process exiting.
    
        Returns ``True`` if all processes have been joined successfully,
        ``False`` if there are more processes that need to be joined.
    
        Args:
            timeout (float): Wait this long (in seconds) before giving up on waiting.
            grace_period (float): When any processes fail, wait this long (in seconds)
                for others to shutdown gracefully before terminating them. If they
                still don't exit, wait another grace period before killing them.
        """
        # Ensure this function can be called even when we're done.
        if len(self.sentinels) == 0:
            return True
    
        # Wait for any process to fail or all of them to succeed.
        ready = multiprocessing.connection.wait(
            self.sentinels.keys(),
            timeout=timeout,
        )
    
        error_index = None
        for sentinel in ready:
            index = self.sentinels.pop(sentinel)
            process = self.processes[index]
            process.join()
            if process.exitcode != 0:
                error_index = index
                break
    
        # Return if there was no error.
        if error_index is None:
            # Return whether or not all processes have been joined.
            return len(self.sentinels) == 0
        # An error occurred. Clean-up all processes before returning.
        # First, allow a grace period for processes to shutdown themselves.
        if grace_period is not None:
            self._join_procs_with_timeout(grace_period)
        # Then, terminate processes that are still alive. Try SIGTERM first.
        for process in self.processes:
            if process.is_alive():
                log.warning("Terminating process %s via signal SIGTERM", process.pid)
                process.terminate()
    
        # Try SIGKILL if the process isn't going down after another grace_period.
        # The reason is related to python signal handling is limited
        # to main thread and if that is in c/c++ land and stuck it won't
        # to handle it. We have seen processes getting stuck not handling
        # SIGTERM for the above reason.
        self._join_procs_with_timeout(30 if grace_period is None else grace_period)
        for process in self.processes:
            if process.is_alive():
                log.warning(
                    "Unable to shutdown process %s via SIGTERM , forcefully exiting via SIGKILL",
                    process.pid,
                )
                process.kill()
            process.join()
    
        # The file will only be created if the process crashed.
        failed_process = self.processes[error_index]
        if not os.access(self.error_files[error_index], os.R_OK):
            exitcode = self.processes[error_index].exitcode
            if exitcode < 0:
                try:
                    name = signal.Signals(-exitcode).name
                except ValueError:
                    name = f"<Unknown signal {-exitcode}>"
>               raise ProcessExitedException(
                    f"process {error_index:d} terminated with signal {name}",
                    error_index=error_index,
                    error_pid=failed_process.pid,
                    exit_code=exitcode,
                    signal_name=name,
                )
E               torch.multiprocessing.spawn.ProcessExitedException: process 1 terminated with signal SIGABRT

../../.venv/lib/python3.12.../torch/multiprocessing/spawn.py:192: ProcessExitedException

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}