Add Support for Non-primitive types in TensorboardLogger

CHANGELOG.md

@@ -10,6 +10,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 - Added type hints to `pytorch_lightning.core` ([#946](https://github.com/PyTorchLightning/pytorch-lightning/pull/946))
 - Added support for IterableDataset in validation and testing ([#1104](https://github.com/PyTorchLightning/pytorch-lightning/pull/1104))
+- Added support for non-primitive types in hparams for TensorboardLogger ([#1130](https://github.com/PyTorchLightning/pytorch-lightning/pull/1130))
+
 
 ### Changed
 

pytorch_lightning/loggers/base.py

@@ -4,6 +4,8 @@
 from functools import wraps
 from typing import Union, Optional, Dict, Iterable, Any, Callable, List
 
+import torch
+
 
 def rank_zero_only(fn: Callable):
     """Decorate a logger method to run it only on the process with rank 0.
@@ -42,7 +44,8 @@ def log_metrics(self, metrics: Dict[str, float], step: Optional[int] = None):
         """
         pass
 
-    def _convert_params(self, params: Union[Dict[str, Any], Namespace]) -> Dict[str, Any]:
+    @staticmethod
+    def _convert_params(params: Union[Dict[str, Any], Namespace]) -> Dict[str, Any]:
         # in case converting from namespace
         if isinstance(params, Namespace):
             params = vars(params)
@@ -52,6 +55,29 @@ def _convert_params(self, params: Union[Dict[str, Any], Namespace]) -> Dict[str,
 
         return params
 
+    @staticmethod
+    def _sanitize_params(params: Dict[str, Any]) -> Dict[str, Any]:
+        """Returns params with non-primitvies converted to strings for logging
+
+        >>> params = {"float": 0.3,
+        ...           "int": 1,
+        ...           "string": "abc",
+        ...           "bool": True,
+        ...           "list": [1, 2, 3],
+        ...           "namespace": Namespace(foo=3),
+        ...           "layer": torch.nn.BatchNorm1d}
+        >>> import pprint
+        >>> pprint.pprint(LightningLoggerBase._sanitize_params(params))  # doctest: +NORMALIZE_WHITESPACE
+        {'bool': True,
+         'float': 0.3,
+         'int': 1,
+         'layer': "<class 'torch.nn.modules.batchnorm.BatchNorm1d'>",
+         'list': '[1, 2, 3]',
+         'namespace': 'Namespace(foo=3)',
+         'string': 'abc'}
+        """
+        return {k: v if type(v) in [bool, int, float, str, torch.Tensor] else str(v) for k, v in params.items()}
+
     @abstractmethod
     def log_hyperparams(self, params: argparse.Namespace):
         """Record hyperparameters.

pytorch_lightning/loggers/tensorboard.py

@@ -101,6 +101,7 @@ def experiment(self) -> SummaryWriter:
     @rank_zero_only
     def log_hyperparams(self, params: Union[Dict[str, Any], Namespace]) -> None:
         params = self._convert_params(params)
+        sanitized_params = self._sanitize_params(params)
 
         if parse_version(torch.__version__) < parse_version("1.3.0"):
             warn(
@@ -110,13 +111,14 @@ def log_hyperparams(self, params: Union[Dict[str, Any], Namespace]) -> None:
             )
         else:
             from torch.utils.tensorboard.summary import hparams
-            exp, ssi, sei = hparams(params, {})
+            exp, ssi, sei = hparams(sanitized_params, {})
             writer = self.experiment._get_file_writer()
             writer.add_summary(exp)
             writer.add_summary(ssi)
             writer.add_summary(sei)
+
         # some alternative should be added
-        self.tags.update(params)
+        self.tags.update(sanitized_params)
 
     @rank_zero_only
     def log_metrics(self, metrics: Dict[str, float], step: Optional[int] = None) -> None:

tests/loggers/test_tensorboard.py

@@ -1,4 +1,5 @@
 import pickle
+from argparse import Namespace
 
 import pytest
 import torch
@@ -108,6 +109,9 @@ def test_tensorboard_log_hyperparams(tmpdir):
         "float": 0.3,
         "int": 1,
         "string": "abc",
-        "bool": True
+        "bool": True,
+        "list": [1, 2, 3],
+        "namespace": Namespace(foo=3),
+        "layer": torch.nn.BatchNorm1d
     }
     logger.log_hyperparams(hparams)