fix warnings and failures

dcgan/main.py

@@ -224,7 +224,8 @@ def forward(self, input):
         netD.zero_grad()
         real_cpu = data[0].to(device)
         batch_size = real_cpu.size(0)
-        label = torch.full((batch_size,), real_label, device=device)
+        label = torch.full((batch_size,), real_label,
+                           dtype=real_cpu.dtype, device=device)
 
         output = netD(real_cpu)
         errD_real = criterion(output, label)

fast_neural_style/download_saved_models.py

@@ -12,7 +12,10 @@
 try:
     from torch.utils.model_zoo import _download_url_to_file
 except ImportError:
-    from torch.hub import _download_url_to_file
+    try:
+        from torch.hub import download_url_to_file as _download_url_to_file
+    except ImportError:
+        from torch.hub import _download_url_to_file
 
 
 def unzip(source_filename, dest_dir):

run_python_examples.sh

@@ -148,6 +148,7 @@ function word_language_model() {
 
 function clean() {
   cd $BASE_DIR
+  echo "running clean to remove cruft"
   rm -rf dcgan/_cache_lsun_classroom_train_lmdb \
     dcgan/fake_samples_epoch_000.png dcgan/lsun/ \
     dcgan/_cache_lsunclassroomtrainlmdb \
@@ -194,17 +195,20 @@ if [ "" == "$EXAMPLES" ]; then
 else
   for i in $(echo $EXAMPLES | sed "s/,/ /g")
   do
+    echo "Starting $i"
     $i
+    echo "Finished $i, status $?"
   done
 fi
 
 if [ "" == "$ERRORS" ]; then
-  tput setaf 2
-  echo "Completed successfully"
+  [[ "$TERM" != "" ]] && [[ "$TERM" != "dumb" ]]  && tput setaf 2
+  echo "Completed successfully with status $?"
 else
-  tput setaf 1
+  [[ "$TERM" != "" ]] && [[ "$TERM" != "dumb" ]]  && tput setaf 1
   echo "Some examples failed:"
   printf "$ERRORS"
 fi
 
-tput sgr0
+[[ "$TERM" != "" ]] && [[ "$TERM" != "dumb" ]]  && tput sgr0
+

word_language_model/main.py

@@ -180,7 +180,7 @@ def train():
         # `clip_grad_norm` helps prevent the exploding gradient problem in RNNs / LSTMs.
         torch.nn.utils.clip_grad_norm_(model.parameters(), args.clip)
         for p in model.parameters():
-            p.data.add_(-lr, p.grad)
+            p.data.add_(p.grad, alpha=-lr)
 
         total_loss += loss.item()