check gpu availability to avoid confusion

examples/DeepQNetwork/atari.py

@@ -54,7 +54,7 @@ class AtariPlayer(RLEnvironment):
             "rom {} not found. Please download at {}".format(rom_file, ROM_URL)
 
         try:
-            ALEInterface.setLoggerMode(ALEInterface.Logger.Warning)
+            ALEInterface.setLoggerMode(ALEInterface.Logger.Error)
         except AttributeError:
             if execute_only_once():
                 logger.warn("You're not using latest ALE")

examples/GAN/DiscoGAN-CelebA.py

@@ -211,6 +211,7 @@ if __name__ == '__main__':
     parser.add_argument('--load', help='load model')
     args = parser.parse_args()
 
+    assert tf.test.is_gpu_available()
     logger.auto_set_dir()
 
     data = get_celebA_data(args.data, args.style_A, args.style_B)

examples/ResNet/cifar10-resnet.py

@@ -47,6 +47,7 @@ class Model(ModelDesc):
     def _build_graph(self, inputs):
         image, label = inputs
         image = image / 128.0
+        assert tf.test.is_gpu_available()
         image = tf.transpose(image, [0, 3, 1, 2])
 
         def residual(name, l, increase_dim=False, first=False):

examples/ResNet/imagenet-resnet.py

@@ -23,6 +23,8 @@ DEPTH = None
 
 class Model(ModelDesc):
     def __init__(self, data_format='NCHW'):
+        if data_format == 'NCHW':
+            assert tf.test.is_gpu_available()
         self.data_format = data_format
 
     def _get_inputs(self):

examples/cifar-convnet.py

@@ -40,11 +40,15 @@ class Model(ModelDesc):
 
         if is_training:
             tf.summary.image("train_image", image, 10)
-        image = tf.transpose(image, [0, 3, 1, 2])
+        if tf.test.is_gpu_available():
+            image = tf.transpose(image, [0, 3, 1, 2])
+            data_format = 'NCHW'
+        else:
+            data_format = 'NHWC'
 
         image = image / 4.0     # just to make range smaller
         with argscope(Conv2D, nl=BNReLU, use_bias=False, kernel_shape=3), \
-                argscope([Conv2D, MaxPooling, BatchNorm], data_format='NCHW'):
+                argscope([Conv2D, MaxPooling, BatchNorm], data_format=data_format):
             logits = LinearWrap(image) \
                 .Conv2D('conv1.1', out_channel=64) \
                 .Conv2D('conv1.2', out_channel=64) \

tensorpack/RL/history.py

@@ -5,11 +5,38 @@
 
 import numpy as np
 from collections import deque
+from six.moves import range
 from .envbase import ProxyPlayer
 
 __all__ = ['HistoryFramePlayer']
 
 
+class HistoryBuffer(object):
+    def __init__(self, hist_len, concat_axis=2):
+        self.buf = deque(maxlen=hist_len)
+        self.concat_axis = concat_axis
+
+    def push(self, s):
+        self.buf.append(s)
+
+    def clear(self):
+        self.buf.clear()
+
+    def get(self):
+        difflen = self.buf.maxlen - len(self.buf)
+        if difflen == 0:
+            ret = self.buf
+        else:
+            zeros = [np.zeros_like(self.buf[0]) for k in range(difflen)]
+            for k in self.buf:
+                zeros.append(k)
+            ret = zeros
+        return np.concatenate(ret, axis=self.concat_axis)
+
+    def __len__(self):
+        return len(self.buf)
+
+
 class HistoryFramePlayer(ProxyPlayer):
     """ Include history frames in state, or use black images.
         It assumes the underlying player will do auto-restart.
@@ -22,30 +49,23 @@ class HistoryFramePlayer(ProxyPlayer):
                 and `hist_len-1` history.
         """
         super(HistoryFramePlayer, self).__init__(player)
-        self.history = deque(maxlen=hist_len)
+        self.history = HistoryBuffer(hist_len, 2)
 
         s = self.player.current_state()
-        self.history.append(s)
+        self.history.push(s)
 
     def current_state(self):
         assert len(self.history) != 0
-        diff_len = self.history.maxlen - len(self.history)
-        if diff_len == 0:
-            return np.concatenate(self.history, axis=2)
-        zeros = [np.zeros_like(self.history[0]) for k in range(diff_len)]
-        for k in self.history:
-            zeros.append(k)
-        assert len(zeros) == self.history.maxlen
-        return np.concatenate(zeros, axis=2)
+        return self.history.get()
 
     def action(self, act):
         r, isOver = self.player.action(act)
         s = self.player.current_state()
-        self.history.append(s)
+        self.history.push(s)
 
         if isOver:  # s would be a new episode
             self.history.clear()
-            self.history.append(s)
+            self.history.push(s)
         return (r, isOver)
 
     def restart_episode(self):