Merge branch 'main' into release-1.2.0

.github/workflows/pypi_release.yml

@@ -35,7 +35,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
 
     steps:
       - uses: actions/checkout@v4

.github/workflows/test.yml

@@ -41,7 +41,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
 
     steps:
       - uses: actions/checkout@v4
@@ -80,7 +80,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.8", "3.9", "3.10", "3.11", "3.12"]
+        python-version: ["3.9", "3.10", "3.11", "3.12"]
 
     steps:
       - uses: actions/checkout@v4

CHANGELOG.md

@@ -75,14 +75,14 @@ Maintenance update widening compatibility ranges and consolidating dependencies:
 
 ### Dependency changes
 
-* the following are no longer core dependencies and have been changed to optional dependencies : `optuna`, `statsmodels`, `pytorch-optimize`, `matplotlib`. Environments relying on functionality requiring these dependencies need to be updated to instlal these explicitly.
+* the following are no longer core dependencies and have been changed to optional dependencies : `optuna`, `statsmodels`, `pytorch-optimize`, `matplotlib`. Environments relying on functionality requiring these dependencies need to be updated to install these explicitly.
 * `optuna` bounds have been updated to `optuna >=3.1.0,<4.0.0`
 * `optuna-integrate` is now an additional soft dependency, in case of `optuna >=3.3.0`
 
 ### Deprecations and removals
 
 * from 1.2.0, the default optimizer will be changed from `"ranger"` to `"adam"` to avoid non-`torch` dependencies in defaults. `pytorch-optimize` optimizers can still be used. Users should set the optimizer explicitly to continue using `"ranger"`.
-*  from 1.1.0, the loggers do not log figures if soft dependency `matplotlib` is not present, but will raise no exceptions in this case. To log figures, ensure tha `matplotlib` is installed.
+*  from 1.1.0, the loggers do not log figures if soft dependency `matplotlib` is not present, but will raise no exceptions in this case. To log figures, ensure that `matplotlib` is installed.
 
 ## v1.0.0 Update to pytorch 2.0 (10/04/2023)
 

CODEOWNERS

@@ -0,0 +1,3 @@
+# The file specifies framework level core developers for automated review requests
+
+* @benheid @fkiraly @fnhirwa @geetu040 @jdb78 @pranavvp16 @XinyuWuu @yarnabrina

pyproject.toml

@@ -66,12 +66,11 @@ version = "1.2.0"            # is being replaced automatically
 authors = [
   {name = "Jan Beitner"},
 ]
-requires-python = ">=3.8,<3.13"
+requires-python = ">=3.9,<3.13"
 classifiers = [
   "Intended Audience :: Developers",
   "Intended Audience :: Science/Research",
   "Programming Language :: Python :: 3",
-  "Programming Language :: Python :: 3.8",
   "Programming Language :: Python :: 3.9",
   "Programming Language :: Python :: 3.10",
   "Programming Language :: Python :: 3.11",
@@ -115,14 +114,14 @@ dependencies = [
 all_extras = [
   "cpflows",
   "matplotlib",
-  "optuna >=3.1.0,<4.0.0",
+  "optuna >=3.1.0,<5.0.0",
   "optuna-integration",
   "pytorch_optimizer >=2.5.1,<4.0.0",
   "statsmodels",
 ]
 
 tuning = [
-  "optuna >=3.1.0,<4.0.0",
+  "optuna >=3.1.0,<5.0.0",
   "optuna-integration",
   "statsmodels",
 ]

pytorch_forecasting/models/base_model.py

@@ -955,6 +955,18 @@ def log_interval(self) -> float:
         else:
             return self.hparams.log_val_interval
 
+    def _logger_supports(self, method: str) -> bool:
+        """Whether logger supports method.
+
+        Returns
+        -------
+        supports_method : bool
+            True if attribute self.logger.experiment.method exists, False otherwise.
+        """
+        if not hasattr(self, "logger") or not hasattr(self.logger, "experiment"):
+            return False
+        return hasattr(self.logger.experiment, method)
+
     def log_prediction(
         self, x: Dict[str, torch.Tensor], out: Dict[str, torch.Tensor], batch_idx: int, **kwargs
     ) -> None:
@@ -981,6 +993,10 @@ def log_prediction(
             if not mpl_available:
                 return None  # don't log matplotlib plots if not available
 
+            # Don't log figures if add_figure is not available
+            if not self._logger_supports("add_figure"):
+                return None
+
             for idx in log_indices:
                 fig = self.plot_prediction(x, out, idx=idx, add_loss_to_title=True, **kwargs)
                 tag = f"{self.current_stage} prediction"
@@ -1156,7 +1172,8 @@ def log_gradient_flow(self, named_parameters: Dict[str, torch.Tensor]) -> None:
 
         mpl_available = _check_matplotlib("log_gradient_flow", raise_error=False)
 
-        if not mpl_available:
+        # Don't log figures if matplotlib or add_figure is not available
+        if not mpl_available or not self._logger_supports("add_figure"):
             return None
 
         import matplotlib.pyplot as plt

pytorch_forecasting/models/nbeats/__init__.py

@@ -277,7 +277,8 @@ def log_interpretation(self, x, out, batch_idx):
         """
         mpl_available = _check_matplotlib("log_interpretation", raise_error=False)
 
-        if not mpl_available:
+        # Don't log figures if matplotlib or add_figure is not available
+        if not mpl_available or not self._logger_supports("add_figure"):
             return None
 
         label = ["val", "train"][self.training]

pytorch_forecasting/models/nhits/__init__.py

@@ -567,7 +567,8 @@ def log_interpretation(self, x, out, batch_idx):
         """
         mpl_available = _check_matplotlib("log_interpretation", raise_error=False)
 
-        if not mpl_available:
+        # Don't log figures if matplotlib or add_figure is not available
+        if not mpl_available or not self._logger_supports("add_figure"):
             return None
 
         label = ["val", "train"][self.training]

pytorch_forecasting/models/temporal_fusion_transformer/__init__.py

@@ -846,7 +846,8 @@ def log_interpretation(self, outputs):
 
         mpl_available = _check_matplotlib("log_interpretation", raise_error=False)
 
-        if not mpl_available:
+        # Don't log figures if matplotlib or add_figure is not available
+        if not mpl_available or not self._logger_supports("add_figure"):
             return None
 
         import matplotlib.pyplot as plt
@@ -885,6 +886,11 @@ def log_embeddings(self):
         """
         Log embeddings to tensorboard
         """
+
+        # Don't log embeddings if add_embedding is not available
+        if not self._logger_supports("add_embedding"):
+            return None
+
         for name, emb in self.input_embeddings.items():
             labels = self.hparams.embedding_labels[name]
             self.logger.experiment.add_embedding(