=> Creating model from file: models/densenet.lua	
nn.Sequential {
  [input -> (1) -> (2) -> (3) -> (4) -> (5) -> (6) -> (7) -> (8) -> (9) -> (10) -> (11) -> (12) -> (13) -> (14) -> (15) -> (16) -> (17) -> (18) -> (19) -> (20) -> (21) -> (22) -> (23) -> (24) -> (25) -> (26) -> (27) -> (28) -> (29) -> (30) -> (31) -> (32) -> (33) -> (34) -> (35) -> (36) -> (37) -> (38) -> (39) -> (40) -> (41) -> (42) -> (43) -> (44) -> (45) -> (46) -> (47) -> (48) -> (49) -> (50) -> output]
  (1): cudnn.SpatialConvolution(3 -> 24, 3x3, 1,1, 1,1) without bias
  (2): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(24 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (3): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(36 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (4): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(48 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (5): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(60 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (6): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(72 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (7): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(84 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (8): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(96 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (9): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(108 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (10): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(120 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (11): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(132 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (12): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(144 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (13): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(156 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (14): cudnn.SpatialBatchNormalization
  (15): cudnn.ReLU
  (16): cudnn.SpatialConvolution(168 -> 84, 1x1) without bias
  (17): cudnn.SpatialAveragePooling(2x2, 2,2)
  (18): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(84 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (19): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(96 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (20): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(108 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (21): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(120 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (22): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(132 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (23): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(144 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (24): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(156 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (25): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(168 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (26): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(180 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (27): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(192 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (28): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(204 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (29): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(216 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (30): cudnn.SpatialBatchNormalization
  (31): cudnn.ReLU
  (32): cudnn.SpatialConvolution(228 -> 114, 1x1) without bias
  (33): cudnn.SpatialAveragePooling(2x2, 2,2)
  (34): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(114 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (35): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(126 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (36): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(138 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (37): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(150 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (38): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(162 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (39): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(174 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (40): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(186 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (41): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(198 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (42): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(210 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (43): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(222 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (44): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(234 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (45): nn.Sequential {
    [input -> (1) -> output]
    (1): nn.Concat {
      input
        |`-> (1): nn.Identity
         `-> (2): nn.Sequential {
               [input -> (1) -> (2) -> (3) -> output]
               (1): cudnn.SpatialBatchNormalization
               (2): cudnn.ReLU
               (3): cudnn.SpatialConvolution(246 -> 12, 3x3, 1,1, 1,1) without bias
             }
         ... -> output
    }
  }
  (46): cudnn.SpatialBatchNormalization
  (47): cudnn.ReLU
  (48): cudnn.SpatialAveragePooling(8x8, 8,8)
  (49): nn.Reshape(258)
  (50): nn.Linear(258 -> 10)
}
 599050
[torch.LongStorage of size 1]