typing: Use Self from typing_extensions (#2494)

Summary:
<!--
Thank you for sending the PR! We appreciate you spending the time to make BoTorch better.

Help us understand your motivation by explaining why you decided to make this change.

You can learn more about contributing to BoTorch here: https://github.com/pytorch/botorch/blob/main/CONTRIBUTING.md
-->

## Motivation
Why wait till 3.11 :) The dependancy of `typing_extensions` is already present in `torch` and is a core python development library with the same license as python itself.

### Have you read the [Contributing Guidelines on pull requests](https://github.com/pytorch/botorch/blob/main/CONTRIBUTING.md#pull-requests)?

Yes

Pull Request resolved: #2494

Test Plan:
I believe this only needs to pass the lint checks. However as it adds a new dependancy (which should be installed already with `torch`), there could be issues with action runners if installation of dependencies is done in a non-standard manner. [However it seems like it should be fine.](https://github.com/pytorch/botorch/blob/d52205a031ecd83b9ef73ba05ca990a284179edb/.github/workflows/test.yml#L45)

## Related PRs

* Closes #2487  (issue, not PR)

Reviewed By: saitcakmak

Differential Revision: D61976244

Pulled By: Balandat

fbshipit-source-id: be3dc67147a1bd210d336e56488f3c6c25b67939

botorch/models/approximate_gp.py

@@ -32,7 +32,7 @@
 import copy
 import warnings
 
-from typing import Optional, TypeVar, Union
+from typing import Optional, Union
 
 import torch
 from botorch.exceptions.warnings import UserInputWarning
@@ -68,9 +68,9 @@
 )
 from torch import Tensor
 from torch.nn import Module
+from typing_extensions import Self
 
 
-TApproxModel = TypeVar("TApproxModel", bound="ApproximateGPyTorchModel")
 TRANSFORM_WARNING = (
     "Using an {ttype} transform with `SingleTaskVariationalGP`. If this "
     "model is trained in minibatches, a {ttype} transform with learnable "
@@ -132,11 +132,11 @@ def __init__(
     def num_outputs(self):
         return self._desired_num_outputs
 
-    def eval(self: TApproxModel) -> TApproxModel:
+    def eval(self) -> Self:
         r"""Puts the model in `eval` mode."""
         return Module.eval(self)
 
-    def train(self: TApproxModel, mode: bool = True) -> TApproxModel:
+    def train(self, mode: bool = True) -> Self:
         r"""Put the model in `train` mode.
 
         Args:

botorch/models/model.py

@@ -16,7 +16,7 @@
 from abc import ABC, abstractmethod
 from collections import defaultdict
 from collections.abc import Mapping
-from typing import Any, Callable, Optional, TYPE_CHECKING, TypeVar, Union
+from typing import Any, Callable, Optional, TYPE_CHECKING, Union
 
 import numpy as np
 import torch
@@ -37,12 +37,11 @@
 from gpytorch.likelihoods.gaussian_likelihood import FixedNoiseGaussianLikelihood
 from torch import Tensor
 from torch.nn import Module, ModuleDict, ModuleList
+from typing_extensions import Self
 
 if TYPE_CHECKING:
     from botorch.acquisition.objective import PosteriorTransform  # pragma: no cover
 
-TFantasizeMixin = TypeVar("TFantasizeMixin", bound="FantasizeMixin")
-
 
 class Model(Module, ABC):
     r"""Abstract base class for BoTorch models.
@@ -289,11 +288,7 @@ def __init__(self, args):
     """
 
     @abstractmethod
-    def condition_on_observations(
-        self: TFantasizeMixin,
-        X: Tensor,
-        Y: Tensor,
-    ) -> TFantasizeMixin:
+    def condition_on_observations(self, X: Tensor, Y: Tensor) -> Self:
         """
         Classes that inherit from `FantasizeMixin` must implement
         a `condition_on_observations` method.
@@ -322,16 +317,13 @@ def transform_inputs(
         a `transform_inputs` method.
         """
 
-    # When Python 3.11 arrives we can start annotating return types like
-    # this as
-    # 'Self', but at this point the verbose 'T...' syntax is needed.
     def fantasize(
-        self: TFantasizeMixin,
+        self,
         X: Tensor,
         sampler: MCSampler,
         observation_noise: Optional[Tensor] = None,
         **kwargs: Any,
-    ) -> TFantasizeMixin:
+    ) -> Self:
         r"""Construct a fantasy model.
 
         Constructs a fantasy model in the following fashion:

requirements.txt

@@ -3,5 +3,6 @@ scipy
 mpmath>=0.19,<=1.3
 torch>=2.0.1
 pyro-ppl>=1.8.4
+typing_extensions
 gpytorch==1.13
 linear_operator==0.5.3