Example/gan none loss (#274)

* Add none loss option for torchbearer models

* Fix error in docs

* Change loss_criterion -> criterion

* Update changelog

CHANGELOG.md

@@ -8,6 +8,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 - Added a verbose level (options are now 0,1,2) which will print progress for the entire fit call, updating every epoch. Useful when doing dynamic programming with little data.
 ### Changed
 - Timer callback can now also be used as a metric which allows display of specified timings to printers and has been moved to metrics.
+- The loss_criterion is renamed to criterion in `torchbearer.Model` arguments.
+- The criterion in `torchbearer.Model` is now optional and will provide a zero loss tensor if it is not given.
 ### Deprecated
 ### Removed
 ### Fixed

docs/_static/examples/gan.py

@@ -94,7 +94,7 @@ def forward(self, real_imgs, state):
         z = Variable(torch.Tensor(np.random.normal(0, 1, (real_imgs.shape[0], latent_dim)))).to(state[tb.DEVICE])
         state[GEN_IMGS] = self.generator(z)
         state[DISC_GEN] = self.discriminator(state[GEN_IMGS])
-        # We don't want to keep discriminator gradients on the generator forward pass
+        # This clears the function graph built up for the discriminator
         self.discriminator.zero_grad()
 
         # Discriminator Forward
@@ -156,10 +156,6 @@ def process(self, state):
         return state[D_LOSS]
 
 
-def zero_loss(y_pred, y_true):
-    return torch.zeros(y_true.shape[0], 1)
-
-
-torchbearermodel = tb.Model(model, optim, zero_loss, ['loss', g_loss(), d_loss()])
+torchbearermodel = tb.Model(model, optim, loss_criterion=None, metrics=['loss', g_loss(), d_loss()])
 torchbearermodel.to(device)
 torchbearermodel.fit_generator(dataloader, epochs=200, pass_state=True, callbacks=[loss_callback, saver_callback])

docs/examples/gan.rst

@@ -43,12 +43,7 @@ Loss
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Since our loss is complicated in this example, we shall forgo the basic loss criterion used in normal torchbearer models.
-
-.. literalinclude:: /_static/examples/gan.py
-   :language: python
-   :lines: 159-160
-
-Instead use a callback to provide the loss. Since this callback is very simple we can use callback decorators on a function (which takes state) to tell torchbearer when it should be called.
+Instead we use a callback to provide the loss. Since this callback is very simple we can use callback decorators on a function (which takes state) to tell torchbearer when it should be called.
 
 .. literalinclude:: /_static/examples/gan.py
    :language: python
@@ -73,7 +68,7 @@ We can then train the torchbearer model on the GPU in the standard way.
 
 .. literalinclude:: /_static/examples/gan.py
    :language: python
-   :lines: 163-165
+   :lines: 159-161
 
 Visualising
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

torchbearer/callbacks/__init__.py

@@ -21,10 +21,6 @@
         :members:
         :undoc-members:
 
-..  automodule:: torchbearer.callbacks.timer
-        :members:
-        :undoc-members:
-
 Tensorboard
 ------------------------------------
 

torchbearer/metrics/__init__.py

@@ -35,6 +35,13 @@
 
 ..  automodule:: torchbearer.metrics.roc_auc_score
         :members:
+
+
+Timer
+------------------------------------
+..  automodule:: torchbearer.metrics.timer
+        :members:
+        :undoc-members:
 """
 
 from .metrics import *

torchbearer/torchbearer.py

@@ -13,22 +13,25 @@
 class Model:
     """ Torchbearermodel to wrap base torch model and provide training environment around it
     """
-    def __init__(self, model, optimizer, loss_criterion, metrics=[]):
+    def __init__(self, model, optimizer, criterion=None, metrics=[]):
         """ Create torchbearermodel which wraps a base torchmodel and provides a training environment surrounding it
 
         :param model: The base pytorch model
         :type model: torch.nn.Module
         :param optimizer: The optimizer used for pytorch model weight updates
         :type optimizer: torch.optim.Optimizer
-        :param loss_criterion: The final loss criterion that provides a loss value to the optimizer
-        :type loss_criterion: function
+        :param criterion: The final loss criterion that provides a loss value to the optimizer
+        :type criterion: function or None
         :param metrics: Additional metrics for display and use within callbacks
         :type metrics: list
         """
         super().__init__()
+        if criterion is None:
+            criterion = lambda y_pred, y_true: torch.zeros(y_true.shape[0], device=y_true.device)
+
         self.main_state = {
             torchbearer.MODEL: model,
-            torchbearer.CRITERION: loss_criterion,
+            torchbearer.CRITERION: criterion,
             torchbearer.OPTIMIZER: optimizer,
             torchbearer.DEVICE: 'cpu',
             torchbearer.DATA_TYPE: torch.float32,