maskrcnn-benchmark理解记录——R_50_FPN的module和各层维度
Table of Contents
序、一些记录
一、先看GeneralizedRCNN
A、backbone
1）ResNet
2）FPN #todo
B、rpn
1） anchor_generator:
2）head #todo
3）box_selector_train（inference.py）#todo
4）box_selector_test
C、rois_heads
1）box
2）keypoint
二、module总览
三、配置总览
序、一些记录
1.记录下COCO数据集：主要从复杂的日常场景中截取，图像中的目标通过精确的segmentation进行位置的标定。图像包括80种物体类别，328,000影像和2,500,000个label。主要的是有约 100,000个人的关键点信息的标注。而coco每张图约4~6个人，那就是约2万张有人的样本。
2.nn.Conv2d的功能是：对由多个输入平面组成的输入信号进行二维卷积。（有bias）
Conc2d（in_channels (int)，out_channels (int)，kernel_size (int or tuple), stride (int or tuple, optional),padding (int or tuple, optional),dilation (int or tuple, optional)，bias (bool, optional)，groups：将输入数据分组，通常不用管这个参数.）
(int1, int2)的元组（本质上单个的int就是相同int的(int, int)）。在元组中，第1个参数对应高度维度，第2个参数对应宽度维度。
bias (bool, optional): If True, adds a learnable bias to the output. Default: True(偏差)
2. 1关于 groups 参数
对于 groups 参数，用于分解 inputs 和 outputs 间的关系，分组进行卷积操作.
[1] - groups=1，所有输入进行卷积操作，得到输出.
[2] - groups=2，等价于有两个并列 conv 操作，每个的输入是一半的 input_channels，并输出一半 - 的 output_channels，然后再进行链接.
[3] - groups=in_channels，每个 input channel 被其自己的 filters 进行卷积操作，尺寸为Cout/Cin
当 group=in_channels 且 out_channels = K * in_channels，其中，K 是正整数，此时的操作被称为 depthwise convolution.
groups决定了将原输入in_channels 分为几组，而每组channel重用几次，由out_channels/groups计算得到，这也说明了为什么需要groups能供被out_channels与in_channels整除. - pyotrch_nn.Conv2d中groups参数的理解
3.总览在后面。
一、先看GeneralizedRCNN
有backbone、rpn、rois_heads。
GeneralizedRCNN(
(backbone): Sequential(
(body): ResNet()
(fpn): FPN()
)
(rpn): RPNModule(
(anchor_generator):AnchorGenerator()
(head): RPNHead()
(box_selector_train): RPNPostProcessor()
(box_selector_test): RPNPostProcessor()
)
(roi_heads): CombinedROIHeads(
(box): ROIBoxHead()
(keypoint): ROIKeypointHead()
)
)A、backbone
1）ResNet
(body): ResNet(
(stem): StemWithFixedBatchNorm()
(layer1): Sequential()
(layer2): Sequential()
(layer3): Sequential()
(layer4): Sequential()#Sequential
)每一层开始有个下采样，并进行维度的调节。eg: layer2(上表中的conv3_x)
先 (downsample): Sequential(
(0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False) 把上层的维度256，转为512以便于与此block【1*1,128；3*3,128;1*1,512】的结尾512维度一致，然后进行add。
2）FPN #todo
(fpn): FPN(
(fpn_inner1): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner2): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner3): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner4): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer4): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(top_blocks): LastLevelMaxPool()
)B、rpn
1） anchor_generator:
(anchor_generator): AnchorGenerator(
(cell_anchors): BufferList()
)2）head #todo
(head): RPNHead(
(conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(cls_logits): Conv2d(256, 3, kernel_size=(1, 1), stride=(1, 1)) #
(bbox_pred): Conv2d(256, 12, kernel_size=(1, 1), stride=(1, 1)) # 位置3*4
)3）box_selector_train（inference.py）#todo
(box_selector_train): RPNPostProcessor()4）box_selector_test
(box_selector_test): RPNPostProcessor()C、rois_heads
1）box
#这里就是RCNN Head
(box): ROIBoxHead(
(feature_extractor): FPN2MLPFeatureExtractor()
#7*7*256.配合spatial_scale对conv1-conv2进行ROIAlign，→fc1024→fc1024
(predictor): FPNPredictor() #class（2）、box（4）
(post_processor): PostProcessor() #从一组classification scores, box regression and proposals,
#计算后处理框（post-processed boxes）, 并应用NMS得到最终结果
)#TODO ：modeling/roi_heads/box_head/inference.py →PostProcessor()
(box): ROIBoxHead(
(feature_extractor): FPN2MLPFeatureExtractor(
(pooler): Pooler(
(poolers): ModuleList(
(0): ROIAlign(output_size=(7, 7), spatial_scale=0.25, sampling_ratio=2)
(1): ROIAlign(output_size=(7, 7), spatial_scale=0.125, sampling_ratio=2)
(2): ROIAlign(output_size=(7, 7), spatial_scale=0.0625, sampling_ratio=2)
(3): ROIAlign(output_size=(7, 7), spatial_scale=0.03125, sampling_ratio=2)
)
)
(fc6): Linear(in_features=12544, out_features=1024, bias=True)
(fc7): Linear(in_features=1024, out_features=1024, bias=True)
)
(predictor): FPNPredictor(
(cls_score): Linear(in_features=1024, out_features=2, bias=True)
(bbox_pred): Linear(in_features=1024, out_features=8, bias=True)
)
(post_processor): PostProcessor()
)
2）keypoint
(keypoint): ROIKeypointHead(
(feature_extractor): KeypointRCNNFeatureExtractor()
(predictor): KeypointRCNNPredictor()
(post_processor): KeypointPostProcessor()
)i)feature_extractor 这里就是 #14*14*256.配合spatial_scale对conv1-conv2进行ROIAlign，就是
14*14*256→14*14*512--*8-->14*14*512(8层)→
ii)predictor （s=3,1/2）7*7*17。共17个热图，也就是类别为17。每个热图大小为7*7，并且呢，每个热图只会分出来一类。
iii) modeling/roi_heads/keypoint_head/inference.py→KeypointPostProcessor TODO
(keypoint): ROIKeypointHead(
(feature_extractor): KeypointRCNNFeatureExtractor(
(pooler): Pooler(
(poolers): ModuleList(
(0): ROIAlign(output_size=(14, 14), spatial_scale=0.25, sampling_ratio=2)
(1): ROIAlign(output_size=(14, 14), spatial_scale=0.125, sampling_ratio=2)
(2): ROIAlign(output_size=(14, 14), spatial_scale=0.0625, sampling_ratio=2)
(3): ROIAlign(output_size=(14, 14), spatial_scale=0.03125, sampling_ratio=2)
)
)
(conv_fcn1): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) #same
(conv_fcn2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn4): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn5): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn6): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn7): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn8): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
)
(predictor): KeypointRCNNPredictor(
(kps_score_lowres): ConvTranspose2d(512, 17, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1))
) # 这个是缩小的1/2
(post_processor): KeypointPostProcessor()
)二、module总览
GeneralizedRCNN(
(backbone): Sequential(
(body): ResNet(
(stem): StemWithFixedBatchNorm(
(conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)
(bn1): FrozenBatchNorm2d()
)
(layer1): Sequential(
(0): BottleneckWithFixedBatchNorm(
(downsample): Sequential(
(0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(1): FrozenBatchNorm2d()
)
(conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(1): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(2): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
)
(layer2): Sequential(
(0): BottleneckWithFixedBatchNorm(
(downsample): Sequential(
(0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): FrozenBatchNorm2d()
)
(conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(1): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(2): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(3): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
)
(layer3): Sequential(
(0): BottleneckWithFixedBatchNorm(
(downsample): Sequential(
(0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): FrozenBatchNorm2d()
)
(conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(1): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(2): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(3): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(4): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(5): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
)
(layer4): Sequential(
(0): BottleneckWithFixedBatchNorm(
(downsample): Sequential(
(0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)
(1): FrozenBatchNorm2d()
)
(conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(1): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
(2): BottleneckWithFixedBatchNorm(
(conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn1): FrozenBatchNorm2d()
(conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)
(bn2): FrozenBatchNorm2d()
(conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)
(bn3): FrozenBatchNorm2d()
)
)
)
(fpn): FPN(
(fpn_inner1): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner2): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner3): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(fpn_inner4): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1))
(fpn_layer4): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(top_blocks): LastLevelMaxPool()
)
)
(rpn): RPNModule(
(anchor_generator): AnchorGenerator(
(cell_anchors): BufferList()
)
(head): RPNHead(
(conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(cls_logits): Conv2d(256, 3, kernel_size=(1, 1), stride=(1, 1))
(bbox_pred): Conv2d(256, 12, kernel_size=(1, 1), stride=(1, 1))
)
(box_selector_train): RPNPostProcessor()
(box_selector_test): RPNPostProcessor()
)
(roi_heads): CombinedROIHeads(
(box): ROIBoxHead(
(feature_extractor): FPN2MLPFeatureExtractor(
(pooler): Pooler(
(poolers): ModuleList(
(0): ROIAlign(output_size=(7, 7), spatial_scale=0.25, sampling_ratio=2)
(1): ROIAlign(output_size=(7, 7), spatial_scale=0.125, sampling_ratio=2)
(2): ROIAlign(output_size=(7, 7), spatial_scale=0.0625, sampling_ratio=2)
(3): ROIAlign(output_size=(7, 7), spatial_scale=0.03125, sampling_ratio=2)
)
)
(fc6): Linear(in_features=12544, out_features=1024, bias=True)
(fc7): Linear(in_features=1024, out_features=1024, bias=True)
)
(predictor): FPNPredictor(
(cls_score): Linear(in_features=1024, out_features=2, bias=True)
(bbox_pred): Linear(in_features=1024, out_features=8, bias=True)
)
(post_processor): PostProcessor()
)
(keypoint): ROIKeypointHead(
(feature_extractor): KeypointRCNNFeatureExtractor(
(pooler): Pooler(
(poolers): ModuleList(
(0): ROIAlign(output_size=(14, 14), spatial_scale=0.25, sampling_ratio=2)
(1): ROIAlign(output_size=(14, 14), spatial_scale=0.125, sampling_ratio=2)
(2): ROIAlign(output_size=(14, 14), spatial_scale=0.0625, sampling_ratio=2)
(3): ROIAlign(output_size=(14, 14), spatial_scale=0.03125, sampling_ratio=2)
)
)
(conv_fcn1): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn3): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn4): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn5): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn6): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn7): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_fcn8): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
)
(predictor): KeypointRCNNPredictor(
(kps_score_lowres): ConvTranspose2d(512, 17, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1))
)
(post_processor): KeypointPostProcessor()
)
)
)三、配置总览
2019-03-04 15:44:25,518 maskrcnn_benchmark.utils.checkpoint INFO: Loading checkpoint from catalog://ImageNetPretrained/MSRA/R-50
2019-03-04 15:44:25,518 maskrcnn_benchmark.utils.checkpoint INFO: catalog://ImageNetPretrained/MSRA/R-50 points to https://dl.fbaipublicfiles.com/detectron/ImageNetPretrained/MSRA/R-50.pkl
2019-03-04 15:44:25,521 maskrcnn_benchmark.utils.checkpoint INFO: url https://dl.fbaipublicfiles.com/detectron/ImageNetPretrained/MSRA/R-50.pkl cached in /home/vivian/.torch/models/R-50.pkl
2019-03-04 15:44:25,851 maskrcnn_benchmark.utils.c2_model_loading INFO: Remapping C2 weights
2019-03-04 15:44:25,851 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: conv1_b mapped name: conv1.bias
2019-03-04 15:44:25,851 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: conv1_w mapped name: conv1.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: fc1000_b mapped name: fc1000.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: fc1000_w mapped name: fc1000.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch1_b mapped name: layer1.0.downsample.0.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch1_bn_b mapped name: layer1.0.downsample.1.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch1_bn_s mapped name: layer1.0.downsample.1.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch1_w mapped name: layer1.0.downsample.0.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2a_b mapped name: layer1.0.conv1.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2a_bn_b mapped name: layer1.0.bn1.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2a_bn_s mapped name: layer1.0.bn1.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2a_w mapped name: layer1.0.conv1.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2b_b mapped name: layer1.0.conv2.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2b_bn_b mapped name: layer1.0.bn2.bias
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2b_bn_s mapped name: layer1.0.bn2.weight
2019-03-04 15:44:25,852 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2b_w mapped name: layer1.0.conv2.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2c_b mapped name: layer1.0.conv3.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2c_bn_b mapped name: layer1.0.bn3.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2c_bn_s mapped name: layer1.0.bn3.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_0_branch2c_w mapped name: layer1.0.conv3.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2a_b mapped name: layer1.1.conv1.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2a_bn_b mapped name: layer1.1.bn1.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2a_bn_s mapped name: layer1.1.bn1.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2a_w mapped name: layer1.1.conv1.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2b_b mapped name: layer1.1.conv2.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2b_bn_b mapped name: layer1.1.bn2.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2b_bn_s mapped name: layer1.1.bn2.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2b_w mapped name: layer1.1.conv2.weight
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2c_b mapped name: layer1.1.conv3.bias
2019-03-04 15:44:25,853 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2c_bn_b mapped name: layer1.1.bn3.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2c_bn_s mapped name: layer1.1.bn3.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_1_branch2c_w mapped name: layer1.1.conv3.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2a_b mapped name: layer1.2.conv1.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2a_bn_b mapped name: layer1.2.bn1.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2a_bn_s mapped name: layer1.2.bn1.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2a_w mapped name: layer1.2.conv1.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2b_b mapped name: layer1.2.conv2.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2b_bn_b mapped name: layer1.2.bn2.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2b_bn_s mapped name: layer1.2.bn2.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2b_w mapped name: layer1.2.conv2.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2c_b mapped name: layer1.2.conv3.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2c_bn_b mapped name: layer1.2.bn3.bias
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2c_bn_s mapped name: layer1.2.bn3.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res2_2_branch2c_w mapped name: layer1.2.conv3.weight
2019-03-04 15:44:25,854 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch1_b mapped name: layer2.0.downsample.0.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch1_bn_b mapped name: layer2.0.downsample.1.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch1_bn_s mapped name: layer2.0.downsample.1.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch1_w mapped name: layer2.0.downsample.0.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2a_b mapped name: layer2.0.conv1.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2a_bn_b mapped name: layer2.0.bn1.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2a_bn_s mapped name: layer2.0.bn1.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2a_w mapped name: layer2.0.conv1.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2b_b mapped name: layer2.0.conv2.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2b_bn_b mapped name: layer2.0.bn2.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2b_bn_s mapped name: layer2.0.bn2.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2b_w mapped name: layer2.0.conv2.weight
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2c_b mapped name: layer2.0.conv3.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2c_bn_b mapped name: layer2.0.bn3.bias
2019-03-04 15:44:25,855 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2c_bn_s mapped name: layer2.0.bn3.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_0_branch2c_w mapped name: layer2.0.conv3.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2a_b mapped name: layer2.1.conv1.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2a_bn_b mapped name: layer2.1.bn1.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2a_bn_s mapped name: layer2.1.bn1.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2a_w mapped name: layer2.1.conv1.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2b_b mapped name: layer2.1.conv2.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2b_bn_b mapped name: layer2.1.bn2.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2b_bn_s mapped name: layer2.1.bn2.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2b_w mapped name: layer2.1.conv2.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2c_b mapped name: layer2.1.conv3.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2c_bn_b mapped name: layer2.1.bn3.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2c_bn_s mapped name: layer2.1.bn3.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_1_branch2c_w mapped name: layer2.1.conv3.weight
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2a_b mapped name: layer2.2.conv1.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2a_bn_b mapped name: layer2.2.bn1.bias
2019-03-04 15:44:25,856 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2a_bn_s mapped name: layer2.2.bn1.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2a_w mapped name: layer2.2.conv1.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2b_b mapped name: layer2.2.conv2.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2b_bn_b mapped name: layer2.2.bn2.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2b_bn_s mapped name: layer2.2.bn2.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2b_w mapped name: layer2.2.conv2.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2c_b mapped name: layer2.2.conv3.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2c_bn_b mapped name: layer2.2.bn3.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2c_bn_s mapped name: layer2.2.bn3.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_2_branch2c_w mapped name: layer2.2.conv3.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2a_b mapped name: layer2.3.conv1.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2a_bn_b mapped name: layer2.3.bn1.bias
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2a_bn_s mapped name: layer2.3.bn1.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2a_w mapped name: layer2.3.conv1.weight
2019-03-04 15:44:25,857 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2b_b mapped name: layer2.3.conv2.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loaStemWithFixedBatchNormding INFO: C2 name: res3_3_branch2b_bn_b mapped name: layer2.3.bn2.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2b_bn_s mapped name: layer2.3.bn2.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2b_w mapped name: layer2.3.conv2.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2c_b mapped name: layer2.3.conv3.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2c_bn_b mapped name: layer2.3.bn3.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2c_bn_s mapped name: layer2.3.bn3.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res3_3_branch2c_w mapped name: layer2.3.conv3.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch1_b mapped name: layer3.0.downsample.0.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch1_bn_b mapped name: layer3.0.downsample.1.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch1_bn_s mapped name: layer3.0.downsample.1.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch1_w mapped name: layer3.0.downsample.0.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2a_b mapped name: layer3.0.conv1.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2a_bn_b mapped name: layer3.0.bn1.bias
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2a_bn_s mapped name: layer3.0.bn1.weight
2019-03-04 15:44:25,858 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2a_w mapped name: layer3.0.conv1.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2b_b mapped name: layer3.0.conv2.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2b_bn_b mapped name: layer3.0.bn2.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2b_bn_s mapped name: layer3.0.bn2.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2b_w mapped name: layer3.0.conv2.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2c_b mapped name: layer3.0.conv3.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2c_bn_b mapped name: layer3.0.bn3.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2c_bn_s mapped name: layer3.0.bn3.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_0_branch2c_w mapped name: layer3.0.conv3.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2a_b mapped name: layer3.1.conv1.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2a_bn_b mapped name: layer3.1.bn1.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2a_bn_s mapped name: layer3.1.bn1.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2a_w mapped name: layer3.1.conv1.weight
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2b_b mapped name: layer3.1.conv2.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2b_bn_b mapped name: layer3.1.bn2.bias
2019-03-04 15:44:25,859 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2b_bn_s mapped name: layer3.1.bn2.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2b_w mapped name: layer3.1.conv2.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2c_b mapped name: layer3.1.conv3.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2c_bn_b mapped name: layer3.1.bn3.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2c_bn_s mapped name: layer3.1.bn3.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_1_branch2c_w mapped name: layer3.1.conv3.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2a_b mapped name: layer3.2.conv1.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2a_bn_b mapped name: layer3.2.bn1.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2a_bn_s mapped name: layer3.2.bn1.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2a_w mapped name: layer3.2.conv1.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2b_b mapped name: layer3.2.conv2.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2b_bn_b mapped name: layer3.2.bn2.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2b_bn_s mapped name: layer3.2.bn2.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2b_w mapped name: layer3.2.conv2.weight
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2c_b mapped name: layer3.2.conv3.bias
2019-03-04 15:44:25,860 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2c_bn_b mapped name: layer3.2.bn3.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2c_bn_s mapped name: layer3.2.bn3.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_2_branch2c_w mapped name: layer3.2.conv3.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2a_b mapped name: layer3.3.conv1.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2a_bn_b mapped name: layer3.3.bn1.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2a_bn_s mapped name: layer3.3.bn1.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2a_w mapped name: layer3.3.conv1.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2b_b mapped name: layer3.3.conv2.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2b_bn_b mapped name: layer3.3.bn2.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2b_bn_s mapped name: layer3.3.bn2.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2b_w mapped name: layer3.3.conv2.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2c_b mapped name: layer3.3.conv3.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2c_bn_b mapped name: layer3.3.bn3.bias
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2c_bn_s mapped name: layer3.3.bn3.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_3_branch2c_w mapped name: layer3.3.conv3.weight
2019-03-04 15:44:25,861 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2a_b mapped name: layer3.4.conv1.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2a_bn_b mapped name: layer3.4.bn1.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2a_bn_s mapped name: layer3.4.bn1.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2a_w mapped name: layer3.4.conv1.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2b_b mapped name: layer3.4.conv2.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2b_bn_b mapped name: layer3.4.bn2.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2b_bn_s mapped name: layer3.4.bn2.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2b_w mapped name: layer3.4.conv2.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2c_b mapped name: layer3.4.conv3.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2c_bn_b mapped name: layer3.4.bn3.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2c_bn_s mapped name: layer3.4.bn3.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_4_branch2c_w mapped name: layer3.4.conv3.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2a_b mapped name: layer3.5.conv1.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2a_bn_b mapped name: layer3.5.bn1.bias
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2a_bn_s mapped name: layer3.5.bn1.weight
2019-03-04 15:44:25,862 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2a_w mapped name: layer3.5.conv1.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2b_b mapped name: layer3.5.conv2.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2b_bn_b mapped name: layer3.5.bn2.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2b_bn_s mapped name: layer3.5.bn2.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2b_w mapped name: layer3.5.conv2.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2c_b mapped name: layer3.5.conv3.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2c_bn_b mapped name: layer3.5.bn3.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2c_bn_s mapped name: layer3.5.bn3.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res4_5_branch2c_w mapped name: layer3.5.conv3.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch1_b mapped name: layer4.0.downsample.0.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch1_bn_b mapped name: layer4.0.downsample.1.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch1_bn_s mapped name: layer4.0.downsample.1.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch1_w mapped name: layer4.0.downsample.0.weight
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2a_b mapped name: layer4.0.conv1.bias
2019-03-04 15:44:25,863 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2a_bn_b mapped name: layer4.0.bn1.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2a_bn_s mapped name: layer4.0.bn1.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2a_w mapped name: layer4.0.conv1.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2b_b mapped name: layer4.0.conv2.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2b_bn_b mapped name: layer4.0.bn2.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2b_bn_s mapped name: layer4.0.bn2.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2b_w mapped name: layer4.0.conv2.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2c_b mapped name: layer4.0.conv3.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2c_bn_b mapped name: layer4.0.bn3.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2c_bn_s mapped name: layer4.0.bn3.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_0_branch2c_w mapped name: layer4.0.conv3.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2a_b mapped name: layer4.1.conv1.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2a_bn_b mapped name: layer4.1.bn1.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2a_bn_s mapped name: layer4.1.bn1.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2a_w mapped name: layer4.1.conv1.weight
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2b_b mapped name: layer4.1.conv2.bias
2019-03-04 15:44:25,864 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2b_bn_b mapped name: layer4.1.bn2.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2b_bn_s mapped name: layer4.1.bn2.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2b_w mapped name: layer4.1.conv2.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2c_b mapped name: layer4.1.conv3.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2c_bn_b mapped name: layer4.1.bn3.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2c_bn_s mapped name: layer4.1.bn3.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_1_branch2c_w mapped name: layer4.1.conv3.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2a_b mapped name: layer4.2.conv1.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2a_bn_b mapped name: layer4.2.bn1.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2a_bn_s mapped name: layer4.2.bn1.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2a_w mapped name: layer4.2.conv1.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2b_b mapped name: layer4.2.conv2.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2b_bn_b mapped name: layer4.2.bn2.bias
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2b_bn_s mapped name: layer4.2.bn2.weight
2019-03-04 15:44:25,865 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2b_w mapped name: layer4.2.conv2.weight
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2c_b mapped name: layer4.2.conv3.bias
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2c_bn_b mapped name: layer4.2.bn3.bias
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2c_bn_s mapped name: layer4.2.bn3.weight
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res5_2_branch2c_w mapped name: layer4.2.conv3.weight
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res_conv1_bn_b mapped name: bn1.bias
2019-03-04 15:44:25,866 maskrcnn_benchmark.utils.c2_model_loading INFO: C2 name: res_conv1_bn_s mapped name: bn1.weight
2019-03-04 15:44:25,882 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn1.bias loaded from layer1.0.bn1.bias of shape (64,)
2019-03-04 15:44:25,882 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn1.weight loaded from layer1.0.bn1.weight of shape (64,)
2019-03-04 15:44:25,882 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn2.bias loaded from layer1.0.bn2.bias of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn2.weight loaded from layer1.0.bn2.weight of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn3.bias loaded from layer1.0.bn3.bias of shape (256,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.bn3.weight loaded from layer1.0.bn3.weight of shape (256,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.conv1.weight loaded from layer1.0.conv1.weight of shape (64, 64, 1, 1)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.conv2.weight loaded from layer1.0.conv2.weight of shape (64, 64, 3, 3)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.conv3.weight loaded from layer1.0.conv3.weight of shape (256, 64, 1, 1)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.downsample.0.weight loaded from layer1.0.downsample.0.weight of shape (256, 64, 1, 1)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.downsample.1.bias loaded from layer1.0.downsample.1.bias of shape (256,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.0.downsample.1.weight loaded from layer1.0.downsample.1.weight of shape (256,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn1.bias loaded from layer1.1.bn1.bias of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn1.weight loaded from layer1.1.bn1.weight of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn2.bias loaded from layer1.1.bn2.bias of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn2.weight loaded from layer1.1.bn2.weight of shape (64,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn3.bias loaded from layer1.1.bn3.bias of shape (256,)
2019-03-04 15:44:25,883 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.bn3.weight loaded from layer1.1.bn3.weight of shape (256,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.conv1.weight loaded from layer1.1.conv1.weight of shape (64, 256, 1, 1)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.conv2.weight loaded from layer1.1.conv2.weight of shape (64, 64, 3, 3)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.1.conv3.weight loaded from layer1.1.conv3.weight of shape (256, 64, 1, 1)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn1.bias loaded from layer1.2.bn1.bias of shape (64,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn1.weight loaded from layer1.2.bn1.weight of shape (64,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn2.bias loaded from layer1.2.bn2.bias of shape (64,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn2.weight loaded from layer1.2.bn2.weight of shape (64,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn3.bias loaded from layer1.2.bn3.bias of shape (256,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.bn3.weight loaded from layer1.2.bn3.weight of shape (256,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.conv1.weight loaded from layer1.2.conv1.weight of shape (64, 256, 1, 1)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.conv2.weight loaded from layer1.2.conv2.weight of shape (64, 64, 3, 3)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer1.2.conv3.weight loaded from layer1.2.conv3.weight of shape (256, 64, 1, 1)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn1.bias loaded from layer2.0.bn1.bias of shape (128,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn1.weight loaded from layer2.0.bn1.weight of shape (128,)
2019-03-04 15:44:25,884 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn2.bias loaded from layer2.0.bn2.bias of shape (128,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn2.weight loaded from layer2.0.bn2.weight of shape (128,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn3.bias loaded from layer2.0.bn3.bias of shape (512,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.bn3.weight loaded from layer2.0.bn3.weight of shape (512,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.conv1.weight loaded from layer2.0.conv1.weight of shape (128, 256, 1, 1)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.conv2.weight loaded from layer2.0.conv2.weight of shape (128, 128, 3, 3)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.conv3.weight loaded from layer2.0.conv3.weight of shape (512, 128, 1, 1)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.downsample.0.weight loaded from layer2.0.downsample.0.weight of shape (512, 256, 1, 1)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.downsample.1.bias loaded from layer2.0.downsample.1.bias of shape (512,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.0.downsample.1.weight loaded from layer2.0.downsample.1.weight of shape (512,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn1.bias loaded from layer2.1.bn1.bias of shape (128,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn1.weight loaded from layer2.1.bn1.weight of shape (128,)
2019-03-04 15:44:25,885 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn2.bias loaded from layer2.1.bn2.bias of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn2.weight loaded from layer2.1.bn2.weight of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn3.bias loaded from layer2.1.bn3.bias of shape (512,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.bn3.weight loaded from layer2.1.bn3.weight of shape (512,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.conv1.weight loaded from layer2.1.conv1.weight of shape (128, 512, 1, 1)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.conv2.weight loaded from layer2.1.conv2.weight of shape (128, 128, 3, 3)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.1.conv3.weight loaded from layer2.1.conv3.weight of shape (512, 128, 1, 1)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn1.bias loaded from layer2.2.bn1.bias of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn1.weight loaded from layer2.2.bn1.weight of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn2.bias loaded from layer2.2.bn2.bias of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn2.weight loaded from layer2.2.bn2.weight of shape (128,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn3.bias loaded from layer2.2.bn3.bias of shape (512,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.bn3.weight loaded from layer2.2.bn3.weight of shape (512,)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.conv1.weight loaded from layer2.2.conv1.weight of shape (128, 512, 1, 1)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.conv2.weight loaded from layer2.2.conv2.weight of shape (128, 128, 3, 3)
2019-03-04 15:44:25,886 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.2.conv3.weight loaded from layer2.2.conv3.weight of shape (512, 128, 1, 1)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn1.bias loaded from layer2.3.bn1.bias of shape (128,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn1.weight loaded from layer2.3.bn1.weight of shape (128,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn2.bias loaded from layer2.3.bn2.bias of shape (128,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn2.weight loaded from layer2.3.bn2.weight of shape (128,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn3.bias loaded from layer2.3.bn3.bias of shape (512,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.bn3.weight loaded from layer2.3.bn3.weight of shape (512,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.conv1.weight loaded from layer2.3.conv1.weight of shape (128, 512, 1, 1)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.conv2.weight loaded from layer2.3.conv2.weight of shape (128, 128, 3, 3)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer2.3.conv3.weight loaded from layer2.3.conv3.weight of shape (512, 128, 1, 1)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn1.bias loaded from layer3.0.bn1.bias of shape (256,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn1.weight loaded from layer3.0.bn1.weight of shape (256,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn2.bias loaded from layer3.0.bn2.bias of shape (256,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn2.weight loaded from layer3.0.bn2.weight of shape (256,)
2019-03-04 15:44:25,887 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn3.bias loaded from layer3.0.bn3.bias of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.bn3.weight loaded from layer3.0.bn3.weight of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.conv1.weight loaded from layer3.0.conv1.weight of shape (256, 512, 1, 1)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.conv2.weight loaded from layer3.0.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.conv3.weight loaded from layer3.0.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.downsample.0.weight loaded from layer3.0.downsample.0.weight of shape (1024, 512, 1, 1)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.downsample.1.bias loaded from layer3.0.downsample.1.bias of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.0.downsample.1.weight loaded from layer3.0.downsample.1.weight of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn1.bias loaded from layer3.1.bn1.bias of shape (256,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn1.weight loaded from layer3.1.bn1.weight of shape (256,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn2.bias loaded from layer3.1.bn2.bias of shape (256,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn2.weight loaded from layer3.1.bn2.weight of shape (256,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn3.bias loaded from layer3.1.bn3.bias of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.bn3.weight loaded from layer3.1.bn3.weight of shape (1024,)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.conv1.weight loaded from layer3.1.conv1.weight of shape (256, 1024, 1, 1)
2019-03-04 15:44:25,888 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.conv2.weight loaded from layer3.1.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.1.conv3.weight loaded from layer3.1.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn1.bias loaded from layer3.2.bn1.bias of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn1.weight loaded from layer3.2.bn1.weight of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn2.bias loaded from layer3.2.bn2.bias of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn2.weight loaded from layer3.2.bn2.weight of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn3.bias loaded from layer3.2.bn3.bias of shape (1024,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.bn3.weight loaded from layer3.2.bn3.weight of shape (1024,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.conv1.weight loaded from layer3.2.conv1.weight of shape (256, 1024, 1, 1)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.conv2.weight loaded from layer3.2.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.2.conv3.weight loaded from layer3.2.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn1.bias loaded from layer3.3.bn1.bias of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn1.weight loaded from layer3.3.bn1.weight of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn2.bias loaded from layer3.3.bn2.bias of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn2.weight loaded from layer3.3.bn2.weight of shape (256,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn3.bias loaded from layer3.3.bn3.bias of shape (1024,)
2019-03-04 15:44:25,889 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.bn3.weight loaded from layer3.3.bn3.weight of shape (1024,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.conv1.weight loaded from layer3.3.conv1.weight of shape (256, 1024, 1, 1)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.conv2.weight loaded from layer3.3.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.3.conv3.weight loaded from layer3.3.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn1.bias loaded from layer3.4.bn1.bias of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn1.weight loaded from layer3.4.bn1.weight of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn2.bias loaded from layer3.4.bn2.bias of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn2.weight loaded from layer3.4.bn2.weight of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn3.bias loaded from layer3.4.bn3.bias of shape (1024,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.bn3.weight loaded from layer3.4.bn3.weight of shape (1024,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.conv1.weight loaded from layer3.4.conv1.weight of shape (256, 1024, 1, 1)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.conv2.weight loaded from layer3.4.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.4.conv3.weight loaded from layer3.4.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn1.bias loaded from layer3.5.bn1.bias of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn1.weight loaded from layer3.5.bn1.weight of shape (256,)
2019-03-04 15:44:25,890 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn2.bias loaded from layer3.5.bn2.bias of shape (256,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn2.weight loaded from layer3.5.bn2.weight of shape (256,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn3.bias loaded from layer3.5.bn3.bias of shape (1024,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.bn3.weight loaded from layer3.5.bn3.weight of shape (1024,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.conv1.weight loaded from layer3.5.conv1.weight of shape (256, 1024, 1, 1)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.conv2.weight loaded from layer3.5.conv2.weight of shape (256, 256, 3, 3)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer3.5.conv3.weight loaded from layer3.5.conv3.weight of shape (1024, 256, 1, 1)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn1.bias loaded from layer4.0.bn1.bias of shape (512,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn1.weight loaded from layer4.0.bn1.weight of shape (512,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn2.bias loaded from layer4.0.bn2.bias of shape (512,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn2.weight loaded from layer4.0.bn2.weight of shape (512,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn3.bias loaded from layer4.0.bn3.bias of shape (2048,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.bn3.weight loaded from layer4.0.bn3.weight of shape (2048,)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.conv1.weight loaded from layer4.0.conv1.weight of shape (512, 1024, 1, 1)
2019-03-04 15:44:25,891 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.conv2.weight loaded from layer4.0.conv2.weight of shape (512, 512, 3, 3)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.conv3.weight loaded from layer4.0.conv3.weight of shape (2048, 512, 1, 1)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.downsample.0.weight loaded from layer4.0.downsample.0.weight of shape (2048, 1024, 1, 1)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.downsample.1.bias loaded from layer4.0.downsample.1.bias of shape (2048,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.0.downsample.1.weight loaded from layer4.0.downsample.1.weight of shape (2048,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn1.bias loaded from layer4.1.bn1.bias of shape (512,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn1.weight loaded from layer4.1.bn1.weight of shape (512,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn2.bias loaded from layer4.1.bn2.bias of shape (512,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn2.weight loaded from layer4.1.bn2.weight of shape (512,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn3.bias loaded from layer4.1.bn3.bias of shape (2048,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.bn3.weight loaded from layer4.1.bn3.weight of shape (2048,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.conv1.weight loaded from layer4.1.conv1.weight of shape (512, 2048, 1, 1)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.conv2.weight loaded from layer4.1.conv2.weight of shape (512, 512, 3, 3)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.1.conv3.weight loaded from layer4.1.conv3.weight of shape (2048, 512, 1, 1)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn1.bias loaded from layer4.2.bn1.bias of shape (512,)
2019-03-04 15:44:25,892 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn1.weight loaded from layer4.2.bn1.weight of shape (512,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn2.bias loaded from layer4.2.bn2.bias of shape (512,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn2.weight loaded from layer4.2.bn2.weight of shape (512,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn3.bias loaded from layer4.2.bn3.bias of shape (2048,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.bn3.weight loaded from layer4.2.bn3.weight of shape (2048,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.conv1.weight loaded from layer4.2.conv1.weight of shape (512, 2048, 1, 1)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.conv2.weight loaded from layer4.2.conv2.weight of shape (512, 512, 3, 3)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.layer4.2.conv3.weight loaded from layer4.2.conv3.weight of shape (2048, 512, 1, 1)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.stem.bn1.bias loaded from bn1.bias of shape (64,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.stem.bn1.weight loaded from bn1.weight of shape (64,)
2019-03-04 15:44:25,893 maskrcnn_benchmark.utils.model_serialization INFO: backbone.body.stem.conv1.weight loaded from conv1.weight of shape (64, 3, 7, 7)
2019-03-04 15:44:25,988 maskrcnn_benchmark.data.build WARNING: When using more than one image per GPU you may encounter an out-of-memory (OOM) error if your GPU does not have sufficient memory. If this happens, you can reduce SOLVER.IMS_PER_BATCH (for training) or TEST.IMS_PER_BATCH (for inference). For training, you must also adjust the learning rate and schedule length according to the linear scaling rule. See for example: https://github.com/facebookresearch/Detectron/blob/master/configs/getting_started/tutorial_1gpu_e2e_faster_rcnn_R-50-FPN.yaml#L14
2019-03-04 15:44:36,408 maskrcnn_benchmark.trainer INFO: Start training
2019-03-04 15:45:06,843 maskrcnn_benchmark.trainer INFO: eta: 2:08:19 iter: 20 loss: 8.4845 (8.6096) loss_classifier: 0.0823 (0.1161) loss_box_reg: 0.0063 (0.0144) loss_kp: 7.9658 (7.9423) loss_objectness: 0.3663 (0.4253) loss_rpn_box_reg: 0.0222 (0.1115) time: 1.4842 (1.5217) data: 0.0101 (0.0850) lr: 0.001793 max mem: 3803
2019-03-04 19:02:22,703 maskrcnn_benchmark INFO: Using 1 GPUs
2019-03-04 19:02:22,703 maskrcnn_benchmark INFO: Namespace(config_file='../configs/e2e_keypoint_rcnn_R_50_FPN_debug.yaml', distributed=False, local_rank=0, opts=[], skip_test=False)
2019-03-04 19:02:22,704 maskrcnn_benchmark INFO: Collecting env info (might take some time)
2019-03-04 19:02:26,314 maskrcnn_benchmark INFO:
PyTorch version: 1.0.0
Is debug build: No
CUDA used to build PyTorch: 9.0.176
OS: Ubuntu 14.04.5 LTS
GCC version: (Ubuntu 4.9.4-2ubuntu1~14.04.1) 4.9.4
CMake version: version 3.5.1
Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: Could not collect
GPU models and configuration:
GPU 0: Quadro K4000
GPU 1: Tesla K40c
Nvidia driver version: 384.130
cuDNN version: Probably one of the following:
/usr/local/MATLAB/R2016b/bin/glnxa64/libcudnn.so.4.0.7
/usr/local/cuda-8.0/lib64/libcudnn.so.6.0.21
/usr/local/cuda-8.0/lib64/libcudnn_static.a
/usr/local/cuda8-cudnn7/lib64/libcudnn.so.7.2.1
/usr/local/cuda8-cudnn7/lib64/libcudnn_static.a
/usr/local/lib/libcudnn.so.5.1.10
/usr/local/lib/libcudnn_static.a
Versions of relevant libraries:
[pip] numpy (1.11.3)
[conda] Could not collect
Pillow (5.4.1)
2019-03-04 19:02:26,315 maskrcnn_benchmark INFO: Loaded configuration file ../configs/e2e_keypoint_rcnn_R_50_FPN_debug.yaml
2019-03-04 19:02:26,315 maskrcnn_benchmark INFO:
MODEL:
META_ARCHITECTURE: "GeneralizedRCNN"
WEIGHT: "catalog://ImageNetPretrained/MSRA/R-50"
BACKBONE:
CONV_BODY: "R-50-FPN"
RESNETS:
BACKBONE_OUT_CHANNELS: 256
RPN:
USE_FPN: True
ANCHOR_STRIDE: (4, 8, 16, 32, 64)
PRE_NMS_TOP_N_TRAIN: 2000 # Per FPN level before nms。RPN 生成proposals的个数，前多少名。然后运用RPN得RPN_POST_NMS_TOP_N
PRE_NMS_TOP_N_TEST: 1000
POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TEST: 1000 #Number of top scoring RPN proposals to keep after combining proposals from
## all FPN levels
ROI_HEADS:
USE_FPN: True
ROI_BOX_HEAD:
POOLER_RESOLUTION: 7
POOLER_SCALES: (0.25, 0.125, 0.0625, 0.03125)
POOLER_SAMPLING_RATIO: 2
FEATURE_EXTRACTOR: "FPN2MLPFeatureExtractor"
PREDICTOR: "FPNPredictor"
NUM_CLASSES: 2
ROI_KEYPOINT_HEAD:
POOLER_SCALES: (0.25, 0.125, 0.0625, 0.03125)
FEATURE_EXTRACTOR: "KeypointRCNNFeatureExtractor"
PREDICTOR: "KeypointRCNNPredictor"
POOLER_RESOLUTION: 14
POOLER_SAMPLING_RATIO: 2
RESOLUTION: 56 # ROI_XFORM_RESOLUTION (14) * UP_SCALE (2) * USE_DECONV_OUTPUT (2)
SHARE_BOX_FEATURE_EXTRACTOR: False
KEYPOINT_ON: True
DATASETS:
TRAIN: ("keypoints_coco_2014_train", "keypoints_coco_2014_valminusminival",)
TEST: ("keypoints_coco_2014_minival",)
INPUT:
MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
DATALOADER:
SIZE_DIVISIBILITY: 32
SOLVER:
BASE_LR: 0.005
WEIGHT_DECAY: 0.0001
STEPS: (5020, 5060)
MAX_ITER: 5080
CHECKPOINT_PERIOD: 5000
IMS_PER_BATCH: 2
OUTPUT_DIR : 'outputs/debug'
TEST:
IMS_PER_BATCH: 1
2019-03-04 19:02:26,316 maskrcnn_benchmark INFO: Running with config:
DATALOADER:
ASPECT_RATIO_GROUPING: True
NUM_WORKERS: 4
SIZE_DIVISIBILITY: 32
DATASETS:
TEST: ('keypoints_coco_2014_minival',)
TRAIN: ('keypoints_coco_2014_train', 'keypoints_coco_2014_valminusminival')
INPUT:
MAX_SIZE_TEST: 1333
MAX_SIZE_TRAIN: 1333
MIN_SIZE_TEST: 800
MIN_SIZE_TRAIN: (640, 672, 704, 736, 768, 800)
PIXEL_MEAN: [102.9801, 115.9465, 122.7717]
PIXEL_STD: [1.0, 1.0, 1.0]
TO_BGR255: True
MODEL:
BACKBONE:
CONV_BODY: R-50-FPN
FREEZE_CONV_BODY_AT: 2
USE_GN: False
CLS_AGNOSTIC_BBOX_REG: False
DEVICE: cuda
FBNET:
ARCH: default
ARCH_DEF:
BN_TYPE: bn
DET_HEAD_BLOCKS: []
DET_HEAD_LAST_SCALE: 1.0
DET_HEAD_STRIDE: 0
DW_CONV_SKIP_BN: True
DW_CONV_SKIP_RELU: True
KPTS_HEAD_BLOCKS: []
KPTS_HEAD_LAST_SCALE: 0.0
KPTS_HEAD_STRIDE: 0
MASK_HEAD_BLOCKS: []
MASK_HEAD_LAST_SCALE: 0.0
MASK_HEAD_STRIDE: 0
RPN_BN_TYPE:
RPN_HEAD_BLOCKS: 0
SCALE_FACTOR: 1.0
WIDTH_DIVISOR: 1
FPN:
USE_GN: False
USE_RELU: False
GROUP_NORM:
DIM_PER_GP: -1
EPSILON: 1e-05
NUM_GROUPS: 32
KEYPOINT_ON: True
MASK_ON: False
META_ARCHITECTURE: GeneralizedRCNN
RESNETS:
BACKBONE_OUT_CHANNELS: 256
NUM_GROUPS: 1
RES2_OUT_CHANNELS: 256
RES5_DILATION: 1
STEM_FUNC: StemWithFixedBatchNorm
STEM_OUT_CHANNELS: 64
STRIDE_IN_1X1: True
TRANS_FUNC: BottleneckWithFixedBatchNorm
WIDTH_PER_GROUP: 64
RETINANET:
ANCHOR_SIZES: (32, 64, 128, 256, 512)
ANCHOR_STRIDES: (8, 16, 32, 64, 128)
ASPECT_RATIOS: (0.5, 1.0, 2.0)
BBOX_REG_BETA: 0.11
BBOX_REG_WEIGHT: 4.0
BG_IOU_THRESHOLD: 0.4
FG_IOU_THRESHOLD: 0.5
INFERENCE_TH: 0.05
LOSS_ALPHA: 0.25
LOSS_GAMMA: 2.0
NMS_TH: 0.4
NUM_CLASSES: 81
NUM_CONVS: 4
OCTAVE: 2.0
PRE_NMS_TOP_N: 1000
PRIOR_PROB: 0.01
SCALES_PER_OCTAVE: 3
STRADDLE_THRESH: 0
USE_C5: True
RETINANET_ON: False
ROI_BOX_HEAD:
CONV_HEAD_DIM: 256
DILATION: 1
FEATURE_EXTRACTOR: FPN2MLPFeatureExtractor
MLP_HEAD_DIM: 1024
NUM_CLASSES: 2
NUM_STACKED_CONVS: 4
POOLER_RESOLUTION: 7
POOLER_SAMPLING_RATIO: 2
POOLER_SCALES: (0.25, 0.125, 0.0625, 0.03125)
PREDICTOR: FPNPredictor
USE_GN: False
ROI_HEADS:
BATCH_SIZE_PER_IMAGE: 512
BBOX_REG_WEIGHTS: (10.0, 10.0, 5.0, 5.0)
BG_IOU_THRESHOLD: 0.5
DETECTIONS_PER_IMG: 100
FG_IOU_THRESHOLD: 0.5
NMS: 0.5
POSITIVE_FRACTION: 0.25
SCORE_THRESH: 0.05
USE_FPN: True
ROI_KEYPOINT_HEAD:
CONV_LAYERS: (512, 512, 512, 512, 512, 512, 512, 512)
FEATURE_EXTRACTOR: KeypointRCNNFeatureExtractor
MLP_HEAD_DIM: 1024
NUM_CLASSES: 17
POOLER_RESOLUTION: 14
POOLER_SAMPLING_RATIO: 2
POOLER_SCALES: (0.25, 0.125, 0.0625, 0.03125)
PREDICTOR: KeypointRCNNPredictor
RESOLUTION: 56
SHARE_BOX_FEATURE_EXTRACTOR: False
ROI_MASK_HEAD:
CONV_LAYERS: (256, 256, 256, 256)
DILATION: 1
FEATURE_EXTRACTOR: ResNet50Conv5ROIFeatureExtractor
MLP_HEAD_DIM: 1024
POOLER_RESOLUTION: 14
POOLER_SAMPLING_RATIO: 0
POOLER_SCALES: (0.0625,)
POSTPROCESS_MASKS: False
POSTPROCESS_MASKS_THRESHOLD: 0.5
PREDICTOR: MaskRCNNC4Predictor
RESOLUTION: 14
SHARE_BOX_FEATURE_EXTRACTOR: True
USE_GN: False
RPN:
ANCHOR_SIZES: (32, 64, 128, 256, 512)
ANCHOR_STRIDE: (4, 8, 16, 32, 64)
ASPECT_RATIOS: (0.5, 1.0, 2.0)
BATCH_SIZE_PER_IMAGE: 256
BG_IOU_THRESHOLD: 0.3
FG_IOU_THRESHOLD: 0.7
FPN_POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TRAIN: 2000
MIN_SIZE: 0
NMS_THRESH: 0.7
POSITIVE_FRACTION: 0.5
POST_NMS_TOP_N_TEST: 1000
POST_NMS_TOP_N_TRAIN: 2000
PRE_NMS_TOP_N_TEST: 1000
PRE_NMS_TOP_N_TRAIN: 2000
RPN_HEAD: SingleConvRPNHead
STRADDLE_THRESH: 0
USE_FPN: True
RPN_ONLY: False
WEIGHT: catalog://ImageNetPretrained/MSRA/R-50
OUTPUT_DIR: outputs/debug
PATHS_CATALOG: /media/hello/helloworld/maskrcnn-keypoint/maskrcnn_benchmark/config/paths_catalog.py
SOLVER:
BASE_LR: 0.005
BIAS_LR_FACTOR: 2
CHECKPOINT_PERIOD: 5000
GAMMA: 0.1
IMS_PER_BATCH: 2
MAX_ITER: 5080
MOMENTUM: 0.9
STEPS: (5020, 5060)
WARMUP_FACTOR: 0.3333333333333333
WARMUP_ITERS: 500
WARMUP_METHOD: linear
WEIGHT_DECAY: 0.0001
WEIGHT_DECAY_BIAS: 0
TEST:
DETECTIONS_PER_IMG: 100
EXPECTED_RESULTS: []
EXPECTED_RESULTS_SIGMA_TOL: 4
IMS_PER_BATCH: 2