faster-rcnn训练自己的数据集出错

fast_rcnn startup done
GPU 1: free memory 7050377625
Use GPU 1
Saving imdb to cache...done
Loading region proposals...done
Warrning: no windows proposal is loaded !
roidb (voc_2007_trainval): 4/32
Saving roidb to cache...done
imdb (voc_2007_test): 1/8
Saving imdb to cache...done
Loading region proposals...done
Warrning: no windows proposal is loaded !
Saving roidb to cache...done
Cleared 0 solvers and 1 stand-alone nets


***************
stage one proposal 
***************
conf:
                      batch_size: 15
                    bg_thresh_hi: 0.3000
                    bg_thresh_lo: 0
                       bg_weight: 1
         drop_boxes_runoff_image: 1
                     feat_stride: 16
                     fg_fraction: 0.5000
                       fg_thresh: 0.7000
                     image_means: [1x1x3 single]
                   ims_per_batch: 1
                        max_size: 600
                        rng_seed: 6
                          scales: 400
                  target_only_gt: 1
                     test_binary: 0
    test_drop_boxes_runoff_image: 0
                   test_max_size: 600
               test_min_box_size: 16
                        test_nms: 0.3000
                     test_scales: 400
                     use_flipped: 1
                         use_gpu: 1
                         anchors: [9x4 double]
                output_width_map: [501x1 containers.Map]
               output_height_map: [501x1 containers.Map]


opts:
           cache_name: 'faster_rcnn_VOC2007_vgg_16layers_stage1_rpn'
                 conf: [1x1 struct]
               do_val: 1
           imdb_train: {[1x1 struct]}
             imdb_val: [1x1 struct]
             net_file: 'D:\faster_rcnn-master vgg\faster_rcnn-master vgg16\faster_rcnn-master vgg16\…'
          roidb_train: {[1x1 struct]}
            roidb_val: [1x1 struct]
    snapshot_interval: 1000
      solver_def_file: 'D:\faster_rcnn-master vgg\faster_rcnn-master vgg16\faster_rcnn-master vgg16\…'
         val_interval: 100
            val_iters: 2


Preparing training data...Starting parallel pool (parpool) using the 'local' profile ... connected to 4 workers.
Done.
Preparing validation data...Done.


------------------------- Iteration 100 -------------------------
Training : err_fg 0.0325, err_bg 0.0771, loss (cls 0.183 + reg 0.0151)
Testing  : err_fg 0, err_bg 0.0714, loss (cls 0.137 + reg 0.0144)


------------------------- Iteration 200 -------------------------
Training : err_fg 0.03, err_bg 0.0771, loss (cls 0.15 + reg 0.0125)
Testing  : err_fg 0.0625, err_bg 0, loss (cls 0.11 + reg 0.00962)


----------------