add imagenet model

LRC/dataset/download_imagenet.sh

+set -e
+if [ "x${IMAGENET_USERNAME}" == x -o "x${IMAGENET_ACCESS_KEY}" == x ];then
+  echo "Please create an account on image-net.org."
+  echo "It will provide you a pair of username and accesskey to download imagenet data."
+  read -p "Username: " IMAGENET_USERNAME
+  read -p "Accesskey: " IMAGENET_ACCESS_KEY
+fi
+
+root_url=http://www.image-net.org/challenges/LSVRC/2012/nnoupb
+valid_tar=ILSVRC2012_img_val.tar
+train_tar=ILSVRC2012_img_train.tar
+train_folder=train/
+valid_folder=val/
+
+echo "Download imagenet training data..."
+mkdir -p ${train_folder}
+wget -nd -c ${root_url}/${train_tar}
+tar xf ${train_tar} -C ${train_folder}
+
+cd ${train_folder}
+for x in `ls *.tar`
+do
+  filename=`basename $x .tar`
+  mkdir -p $filename
+  tar -xf $x -C $filename
+  rm -rf $x
+done
+cd -
+
+echo "Download imagenet validation data..."
+mkdir -p ${valid_folder}
+wget -nd -c ${root_url}/${valid_tar}
+tar xf ${valid_tar} -C ${valid_folder}
+
+echo "Download imagenet label file: val_list.txt & train_list.txt"
+label_file=ImageNet_label.tgz
+label_url=http://imagenet-data.bj.bcebos.com/${label_file}
+wget -nd -c ${label_url}
+tar zxf ${label_file}
+

LRC/model.py

@@ -93,7 +93,8 @@ class Cell():
                         dropout_implementation='upscale_in_train')
             s = h3 + h4
             out += [s]
-        return fluid.layers.concat([out[i] for i in self._concat], axis=1)
+        concat_ = fluid.layers.concat([out[i] for i in self._concat], axis=1, name=name+'concat')
+        return concat_
 
 
 def AuxiliaryHeadCIFAR(input, num_classes, aux_name='auxiliary_head'):
@@ -337,7 +338,7 @@ class NetworkCIFAR(object):
         return lrc_loss_mean
 
 def AuxiliaryHeadImageNet(input, num_classes, aux_name='auxiliary_head'):
-    relu_a = fluid.layers.relu(input, inplace=True)
+    relu_a = fluid.layers.relu(input, inplace=False)
     #relu_a.persistable = True
     #print(relu_a)
     pool_a = fluid.layers.pool2d(relu_a, 5, 'avg', pool_stride=2)
@@ -405,10 +406,11 @@ def Stem0Conv(input, C_out):
         bias_attr=ParamAttr(
             initializer=Constant(0.), name='stem0.1.bias'),
         moving_mean_name='stem0.1.running_mean',
-        moving_variance_name='stem0.1.running_var')
-    relu_a = fluid.layers.relu(bn_a, inplace=True)
+        moving_variance_name='stem0.1.running_var',
+        act='relu')
+    #relu_a = fluid.layers.relu(bn_a,inplace=True)
     conv_b = fluid.layers.conv2d(
-        relu_a,
+        bn_a,
         C_out,
         3,
         padding=1,
@@ -428,7 +430,7 @@ def Stem0Conv(input, C_out):
     return bn_b
 
 def Stem1Conv(input, C_out):
-    relu_a = fluid.layers.relu(input, inplace=True)
+    relu_a = fluid.layers.relu(input,inplace=False)
     conv_a = fluid.layers.conv2d(
         relu_a,
         C_out,

LRC/reader_imagenet.py

@@ -140,7 +140,7 @@ def train(args):
     :rtype: callable
     """
 
-    return reader_creator_filepath(args.data, 'train.txt', True)
+    return reader_creator_filepath(args.data, 'debug.txt', True)
 
 
 def test(args):

LRC/train_imagenet.py

@@ -163,14 +163,19 @@ def train(model, args, im_shape, steps_one_epoch, num_gpu):
     #        return os.path.exists(os.path.join(args.pretrained_model, var.name))
 
     #    fluid.io.load_vars(exe, args.pretrained_model, main_program=train_prog, predicate=if_exist)
-
+    #build_strategy = fluid.BuildStrategy()
+    #build_strategy.enable_inplace = False
+    #build_strategy.memory_optimize = False
+    train_fetch_list = [loss_train]
+  
+    fluid.memory_optimize(train_prog, skip_opt_set=set(train_fetch_list))
     exec_strategy = fluid.ExecutionStrategy()
-    exec_strategy.num_threads = 1
+    #exec_strategy.num_threads = 1
     train_exe = fluid.ParallelExecutor(
-        main_program=train_prog,
-        use_cuda=True,
-        loss_name=loss_train.name,
-        exec_strategy=exec_strategy)
+         main_program=train_prog,
+         use_cuda=True,
+         loss_name=loss_train.name,
+         exec_strategy=exec_strategy)
     
     train_batch_size = args.batch_size
     test_batch_size = 256
@@ -182,8 +187,7 @@ def train(model, args, im_shape, steps_one_epoch, num_gpu):
     test_py_reader.decorate_paddle_reader(test_reader)
 
     fluid.clip.set_gradient_clip(fluid.clip.GradientClipByGlobalNorm(args.grad_clip), program=train_prog)
-    train_fetch_list = [loss_train]
-    fluid.memory_optimize(train_prog, skip_opt_set=set(train_fetch_list))
+    print(train_prog.to_string(True))
 
     def save_model(postfix, main_prog):
         model_path = os.path.join(args.save_model_path, postfix)