clean up docs around loggers (#304)

docs/LightningModule/RequiredTrainerInterface.md

@@ -281,7 +281,7 @@ def validation_step(self, batch, batch_nb):
     # or generated text... or whatever
     sample_imgs = x[:6]
     grid = torchvision.utils.make_grid(sample_imgs)
-    self.experiment.add_image('example_images', grid, 0) 
+    self.logger.experiment.add_image('example_images', grid, 0) 
     
     # calculate acc
     labels_hat = torch.argmax(out, dim=1)

docs/LightningModule/properties.md

@@ -11,18 +11,20 @@ Current dtype
 ---
 #### logger    
 A reference to the logger you passed into trainer. 
+Passing a logger is optional. If you don't pass one in, Lightning will create one for you automatically.
+This logger saves logs to '''/os.getcwd()/lightning_logs'''
 ```python
 Trainer(logger=your_logger)
 ```    
 
 Call it from anywhere in your LightningModule to add metrics, images, etc... whatever your logger supports.  
 
-Here is an example using the Test-tube logger (which is a wrapper on [PyTorch SummaryWriter](https://pytorch.org/docs/stable/tensorboard.html) with versioned folder structure).     
+Here is an example using the TestTubeLogger (which is a wrapper on [PyTorch SummaryWriter](https://pytorch.org/docs/stable/tensorboard.html) with versioned folder structure).     
 ```{.python}
-# if logger is a tensorboard logger or test-tube experiment
-self.logger.add_embedding(...)   
-self.logger.log({'val_loss': 0.9})   
-self.logger.add_scalars(...)   
+# if logger is a tensorboard logger or TestTubeLogger
+self.logger.experiment.add_embedding(...)   
+self.logger.experiment.log({'val_loss': 0.9})   
+self.logger.experiment.add_scalars(...)   
 ```
 
 --- 

docs/Trainer/Logging.md

@@ -81,7 +81,15 @@ If you write a logger than may be useful to others, please send
 a pull request to add it to Lighting!
 
 ---
-### Using loggers
+#### Using loggers
+You can call the logger anywhere from your LightningModule by doing:
+```python
+self.logger
+
+# add an image if using TestTubeLogger
+self.logger.experiment.add_image(...)
+```
+
 
 #### Display metrics in progress bar 
 ``` {.python}

docs/Trainer/hooks.md

@@ -137,5 +137,5 @@ def on_after_backward(self):
         for k, v in params.items():
             grads = v
             name = k
-            self.experiment.add_histogram(tag=name, values=grads, global_step=self.trainer.global_step)
+            self.logger.experiment.add_histogram(tag=name, values=grads, global_step=self.trainer.global_step)
 ```

examples/templates/gan.py

@@ -109,7 +109,7 @@ class GAN(pl.LightningModule):
             # log sampled images
             sample_imgs = self.generated_imgs[:6]
             grid = torchvision.utils.make_grid(sample_imgs)
-            self.experiment.add_image('generated_images', grid, 0)
+            self.logger.experiment.add_image('generated_images', grid, 0)
 
             # ground truth result (ie: all fake)
             valid = torch.ones(imgs.size(0), 1)

pytorch_lightning/root_module/root_module.py

@@ -19,13 +19,14 @@ class LightningModule(GradInformation, ModelIO, ModelHooks):
         self.global_step = 0
         self.loaded_optimizer_states_dict = {}
         self.trainer = None
-        self.experiment = None
+        self.logger = None
         self.example_input_array = None
 
         # track if gpu was requested for checkpointing
         self.on_gpu = False
         self.use_dp = False
         self.use_ddp = False
+        self.use_ddp2 = False
         self.use_amp = False
 
     def forward(self, *args, **kwargs):

pytorch_lightning/trainer/trainer_io.py

@@ -88,7 +88,7 @@ class TrainerIO(object):
         if self.proc_rank == 0:
             # save weights
             print('handling SIGUSR1')
-            self.hpc_save(self.weights_save_path, self.experiment)
+            self.hpc_save(self.weights_save_path, self.logger)
 
             # find job id
             job_id = os.environ['SLURM_JOB_ID']
@@ -105,7 +105,7 @@ class TrainerIO(object):
                 print('requeue failed...')
 
             # close experiment to avoid issues
-            self.experiment.close()
+            self.logger.close()
 
     def term_handler(self, signum, frame):
         # save