Added batch histogram computation (keras-team#6065)

* Added batch histogram computation

* batch_size_histogram renamed to batch_size, default set to 32, added spaces around operators

* PEP8 fix

* Added batch_size in tests/keras/test_callbacks.py::test_TensorBoard_convnet

* PEP8 fix

* Batch size reduced in tests, targets and sample_weights sliced

keras/callbacks.py

@@ -592,6 +592,8 @@ class TensorBoard(Callback):
             write_graph is set to True.
         write_grads: whether to visualize gradient histograms in TensorBoard.
             `histogram_freq` must be greater than 0.
+        batch_size: size of batch of inputs to feed to the network
+            for histograms computation.
         write_images: whether to write model weights to visualize as
             image in TensorBoard.
         embeddings_freq: frequency (in epochs) at which selected embedding
@@ -607,6 +609,7 @@ class TensorBoard(Callback):
 
     def __init__(self, log_dir='./logs',
                  histogram_freq=0,
+                 batch_size=32,
                  write_graph=True,
                  write_grads=False,
                  write_images=False,
@@ -626,6 +629,7 @@ def __init__(self, log_dir='./logs',
         self.embeddings_freq = embeddings_freq
         self.embeddings_layer_names = embeddings_layer_names
         self.embeddings_metadata = embeddings_metadata or {}
+        self.batch_size = batch_size
 
     def set_model(self, model):
         self.model = model
@@ -725,8 +729,6 @@ def on_epoch_end(self, epoch, logs=None):
 
         if self.validation_data and self.histogram_freq:
             if epoch % self.histogram_freq == 0:
-                # TODO: implement batched calls to sess.run
-                # (current call will likely go OOM on GPU)
 
                 val_data = self.validation_data
                 tensors = (self.model.inputs +
@@ -737,10 +739,21 @@ def on_epoch_end(self, epoch, logs=None):
                     tensors += [K.learning_phase()]
 
                 assert len(val_data) == len(tensors)
-                feed_dict = dict(zip(tensors, val_data))
-                result = self.sess.run([self.merged], feed_dict=feed_dict)
-                summary_str = result[0]
-                self.writer.add_summary(summary_str, epoch)
+                val_size = val_data[0].shape[0]
+                i = 0
+                while i < val_size:
+                    step = min(self.batch_size, val_size - i)
+                    batch_val = []
+                    batch_val.append(val_data[0][i:i + step])
+                    batch_val.append(val_data[1][i:i + step])
+                    batch_val.append(val_data[2][i:i + step])
+                    if self.model.uses_learning_phase:
+                        batch_val.append(val_data[3])
+                    feed_dict = dict(zip(tensors, batch_val))
+                    result = self.sess.run([self.merged], feed_dict=feed_dict)
+                    summary_str = result[0]
+                    self.writer.add_summary(summary_str, epoch)
+                    i += self.batch_size
 
         if self.embeddings_freq and self.embeddings_logs:
             if epoch % self.embeddings_freq == 0:

tests/keras/test_callbacks.py

@@ -318,7 +318,8 @@ def data_generator_graph(train):
                   metrics=['accuracy'])
 
     tsb = callbacks.TensorBoard(log_dir=filepath, histogram_freq=1,
-                                write_images=True, write_grads=True)
+                                write_images=True, write_grads=True,
+                                batch_size=5)
     cbks = [tsb]
 
     # fit with validation data
@@ -381,7 +382,8 @@ def test_TensorBoard_convnet():
                   optimizer='rmsprop',
                   metrics=['accuracy'])
     tsb = callbacks.TensorBoard(log_dir=filepath, histogram_freq=1,
-                                write_images=True, write_grads=True)
+                                write_images=True, write_grads=True,
+                                batch_size=16)
     cbks = [tsb]
     model.summary()
     history = model.fit(x_train, y_train, epochs=2, batch_size=16,