remove allow-untyped-defs from ao/quantization/experimental/fake_quantize.py

[bobrenjc93]

Stack from ghstack (oldest at bottom):

• remove allow-untyped-defs from _export/db/logging.py #144093
• remove allow-untyped-defs from torch/mps/event.py #144092
• -> remove allow-untyped-defs from ao/quantization/experimental/fake_quantize.py #144091

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/144091

• 📄 Preview Python docs built from this PR
• 📄 Preview C++ docs built from this PR
• ❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

Note: Links to docs will display an error until the docs builds have been completed.

⏳ No Failures, 3 Pending

As of commit 68d7766 with merge base d88a8c4 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[aorenste]

_P = ParamSpec("_P")
_R = TypeVar("_R")
...

class APoTFakeQuantize(FakeQuantizeBase):
    def __init__(self, observer: Callable[_P, _R] = APoTObserver, *observer_args: _P.args, **observer_kwargs: _P.kwargs):

[aorenste]

(Although that may cascade into a bunch of other type errors...)

[bobrenjc93]

Yeah exactly:

>>> Lint for torch/ao/quantization/experimental/fake_quantize.py:

  Error (MYPY) [assignment]
    Incompatible default for argument "observer" (default has type
    "type[APoTObserver]", argument has type "Callable[_P, _R]")

        20  |    level_indices: Tensor
        21  |
        22  |    def __init__(
    >>> 23  |        self, observer: Callable[_P, _R] = APoTObserver, **observer_kwargs: Any
        24  |    ):
        25  |        super().__init__()
        26  |        self.activation_post_process = observer(**observer_kwargs)

  Error (MYPY) [call-arg]
    Too few arguments

        23  |        self, observer: Callable[_P, _R] = APoTObserver, **observer_kwargs: Any
        24  |    ):
        25  |        super().__init__()
    >>> 26  |        self.activation_post_process = observer(**observer_kwargs)
        27  |        self.dtype = self.activation_post_process.dtype
        28  |
        29  |    def calculate_qparams(self, signed: bool = False) -> tuple[Tensor, Tensor, Tensor, Tensor]:  # type: ignore[override]

  Error (MYPY) [attr-defined]
    "_R" has no attribute "dtype"

        24  |    ):
        25  |        super().__init__()
        26  |        self.activation_post_process = observer(**observer_kwargs)
    >>> 27  |        self.dtype = self.activation_post_process.dtype
        28  |
        29  |    def calculate_qparams(self, signed: bool = False) -> tuple[Tensor, Tensor, Tensor, Tensor]:  # type: ignore[override]
        30  |        return self.activation_post_process.calculate_qparams(signed=signed)

  Error (MYPY) [attr-defined]
    "_R" has no attribute "calculate_qparams"

        27  |        self.dtype = self.activation_post_process.dtype
        28  |
        29  |    def calculate_qparams(self, signed: bool = False) -> tuple[Tensor, Tensor, Tensor, Tensor]:  # type: ignore[override]
    >>> 30  |        return self.activation_post_process.calculate_qparams(signed=signed)
        31  |
        32  |    def forward(self, X: torch.Tensor) -> Tensor:  # type: ignore[override]
        33  |        if self.observer_enabled[0] == 1:

  Error (MYPY) [attr-defined]
    "_R" has no attribute "forward"

        31  |
        32  |    def forward(self, X: torch.Tensor) -> Tensor:  # type: ignore[override]
        33  |        if self.observer_enabled[0] == 1:
    >>> 34  |            self.activation_post_process.forward(X)
        35  |            result = self.activation_post_process.calculate_qparams(signed=False)
        36  |            self.alpha = result[0]
        37  |            self.gamma = result[1]

  Error (MYPY) [attr-defined]
    "_R" has no attribute "calculate_qparams"

        32  |    def forward(self, X: torch.Tensor) -> Tensor:  # type: ignore[override]
        33  |        if self.observer_enabled[0] == 1:
        34  |            self.activation_post_process.forward(X)
    >>> 35  |            result = self.activation_post_process.calculate_qparams(signed=False)
        36  |            self.alpha = result[0]
        37  |            self.gamma = result[1]
        38  |            self.quantization_levels = result[2]

[bobrenjc93]

@pytorchbot merge

[pytorchmergebot]

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here

[cyyever]

@pytorchmergebot merge -f "Unrelated pending"

[pytorchmergebot]

The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command
For more information see pytorch-bot wiki.

[pytorchmergebot]

Merge started

Your change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use -f as last resort and instead consider -i/--ignore-current to continue the merge ignoring current failures. This will allow currently pending tests to finish and report signal before the merge.

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here