unify log_smooth_windows and log_iter (#1530)

* unify log_smooth_windows and log_iter

test=develop

* remove duplicate log_iter

* remain mobile yml, since they are soft-link

test=develop

configs/acfpn/faster_rcnn_r50_vd_acfpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_acfpn_1x/model_final

configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn.yml

 architecture: CornerNetSqueeze
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 snapshot_iter: 10000

configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn_mixup_cosine.yml

 architecture: CornerNetSqueeze
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 snapshot_iter: 10000

configs/anchor_free/cornernet_squeeze_hg104.yml

 architecture: CornerNetSqueeze
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 snapshot_iter: 10000

configs/anchor_free/cornernet_squeeze_r50_vd_fpn.yml

 architecture: CornerNetSqueeze
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 snapshot_iter: 10000

configs/anchor_free/fcos_dcn_r50_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: FCOS
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 5000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/anchor_free/fcos_dcn_r50_fpn_1x_cutmix.yml

@@ -2,7 +2,6 @@ architecture: FCOS
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 5000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/anchor_free/fcos_r50_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: FCOS
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/anchor_free/fcos_r50_fpn_multiscale_2x.yml

@@ -2,7 +2,6 @@ architecture: FCOS
 max_iters: 180000
 use_gpu: true
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/anchor_free/ttfnet_darknet.yml

 architecture: TTFNet
 use_gpu: true
 max_iters: 15000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 1000
 metric: COCO

configs/autoaugment/faster_rcnn_r101_vd_fpn_aa_3x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 270000
 snapshot_iter: 30000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/faster_rcnn_r101_vd_fpn_aa_3x/model_final

configs/autoaugment/faster_rcnn_r50_vd_fpn_aa_3x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 270000
 snapshot_iter: 30000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_aa_3x/model_final

configs/cascade_mask_rcnn_r50_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: CascadeMaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/cascade_rcnn_cls_aware_r101_vd_fpn_1x_softnms.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNNClsAware
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/cascade_rcnn_cls_aware_r101_vd_fpn_1x_softnms/model_final

configs/cascade_rcnn_cls_aware_r101_vd_fpn_ms_test.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNNClsAware
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/cascade_rcnn_cls_aware_r101_vd_fpn_ms_test/model_final

configs/cascade_rcnn_r50_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/cascade_rcnn_r50_fpn_1x/model_final

configs/cascade_rcnn_r50_fpn_1x_ms_test.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/cascade_rcnn_r50_fpn_1x/model_final

configs/dcn/cascade_mask_rcnn_dcnv2_se154_vd_fpn_gn_s1x.yml

@@ -3,7 +3,6 @@ max_iters: 300000
 snapshot_iter: 10
 use_gpu: true
 log_iter: 20
-log_smooth_window: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/SENet154_vd_caffe_pretrained.tar
 weights: output/cascade_mask_rcnn_dcn_se154_vd_fpn_gn_s1x/model_final

configs/dcn/cascade_mask_rcnn_dcnv2_se154_vd_fpn_gn_s1x_ms_test.yml

@@ -3,7 +3,6 @@ max_iters: 300000
 snapshot_iter: 10000
 use_gpu: true
 log_iter: 20
-log_smooth_window: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/SENet154_vd_caffe_pretrained.tar
 weights: output/cascade_mask_rcnn_dcn_se154_vd_fpn_gn_s1x/model_final

configs/dcn/cascade_rcnn_cbr200_vd_fpn_dcnv2_nonlocal_softnms.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNN
 max_iters: 460000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/CBResNet200_vd_pretrained.tar
 weights: output/cascade_rcnn_cbr200_vd_fpn_dcnv2_nonlocal_softnms/model_final

configs/dcn/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNNClsAware
 max_iters: 460000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet200_vd_pretrained.tar
 weights: output/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms/model_final

configs/dcn/cascade_rcnn_dcn_r101_vd_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar

configs/dcn/cascade_rcnn_dcn_r50_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/dcn/cascade_rcnn_dcn_x101_vd_64x4d_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar

configs/dcn/faster_rcnn_dcn_r101_vd_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar

configs/dcn/faster_rcnn_dcn_r50_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: FasterRCNN
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/dcn/faster_rcnn_dcn_r50_vd_fpn_2x.yml

@@ -2,7 +2,6 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar

configs/dcn/faster_rcnn_dcn_x101_vd_64x4d_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar

configs/dcn/mask_rcnn_dcn_r101_vd_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: MaskRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar

configs/dcn/mask_rcnn_dcn_r50_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/dcn/mask_rcnn_dcn_r50_vd_fpn_2x.yml

@@ -2,7 +2,6 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 360000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar

configs/dcn/mask_rcnn_dcn_x101_vd_64x4d_fpn_1x.yml

@@ -2,7 +2,6 @@ architecture: MaskRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar

configs/dcn/yolov3_r50vd_dcn.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 20000
 metric: COCO

configs/dcn/yolov3_r50vd_dcn_db_iouaware_obj365_pretrained_coco.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 85000
-log_smooth_window: 1
+log_iter: 1
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/dcn/yolov3_r50vd_dcn_db_iouloss_obj365_pretrained_coco.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 85000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/dcn/yolov3_r50vd_dcn_db_obj365_pretrained_coco.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 85000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/dcn/yolov3_r50vd_dcn_obj365_pretrained_coco.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 85000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/efficientdet_d0.yml

@@ -3,7 +3,7 @@ max_iters: 281250
 use_gpu: true
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/EfficientNetB0_pretrained.tar
 weights: output/efficientdet_d0/model_final
-log_smooth_window: 20
+log_iter: 20
 snapshot_iter: 10000
 metric: COCO
 save_dir: output

configs/face_detection/blazeface.yml

@@ -3,7 +3,6 @@ max_iters: 320000
 pretrain_weights:
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/face_detection/blazeface_keypoint.yml

@@ -3,7 +3,6 @@ max_iters: 160000
 pretrain_weights:
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/face_detection/blazeface_nas.yml

@@ -3,7 +3,6 @@ max_iters: 320000
 pretrain_weights:
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/face_detection/blazeface_nas_v2.yml

@@ -3,7 +3,6 @@ max_iters: 320000
 pretrain_weights:
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/face_detection/faceboxes.yml

@@ -3,7 +3,6 @@ pretrain_weights:
 use_gpu: true
 max_iters: 320000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/face_detection/faceboxes_lite.yml

@@ -3,7 +3,6 @@ pretrain_weights:
 use_gpu: true
 max_iters: 320000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: output

configs/faster_rcnn_cbr101_vd_dual_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/CBResNet101_vd_pretrained.tar
 weights: output/faster_rcnn_cbr101_vd_dual_fpn_1x/model_final

configs/faster_rcnn_cbr50_vd_dual_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/CBResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_cbr50_vd_dual_fpn_1x/model_final

configs/faster_rcnn_r101_1x.yml

 architecture: FasterRCNN
 use_gpu: true
 max_iters: 180000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.tar

configs/faster_rcnn_r101_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: http://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.tar
 weights: output/faster_rcnn_r101_fpn_1x/model_final

configs/faster_rcnn_r101_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: http://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.tar
 weights: output/faster_rcnn_r101_fpn_2x/model_final

configs/faster_rcnn_r101_vd_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/faster_rcnn_r101_vd_fpn_1x/model_final

configs/faster_rcnn_r101_vd_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/faster_rcnn_r101_vd_fpn_2x/model_final

configs/faster_rcnn_r34_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet34_pretrained.tar
 metric: COCO

configs/faster_rcnn_r34_vd_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet34_vd_pretrained.tar
 metric: COCO

configs/faster_rcnn_r50_1x.yml

 architecture: FasterRCNN
 use_gpu: true
 max_iters: 180000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/faster_rcnn_r50_2x.yml

 architecture: FasterRCNN
 use_gpu: true
 max_iters: 360000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar

configs/faster_rcnn_r50_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/faster_rcnn_r50_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 use_gpu: true
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/faster_rcnn_r50_vd_1x.yml

 architecture: FasterRCNN
 use_gpu: true
 max_iters: 180000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output/faster-r50-vd-c4-1x
 snapshot_iter: 10000
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar

configs/faster_rcnn_r50_vd_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_2x/model_final

configs/faster_rcnn_se154_vd_fpn_s1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 260000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/SENet154_vd_pretrained.tar
 weights: output/faster_rcnn_se154_vd_fpn_s1x/model_final

configs/faster_rcnn_x101_vd_64x4d_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar
 weights: output/faster_rcnn_x101_vd_64x4d_fpn_1x/model_final

configs/faster_rcnn_x101_vd_64x4d_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar
 weights: output/faster_rcnn_x101_vd_64x4d_fpn_1x/model_final

configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 metric: COCO

configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 metric: COCO

configs/gn/cascade_mask_rcnn_r50_fpn_gn_2x.yml

@@ -2,7 +2,7 @@ architecture: CascadeMaskRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/cascade_mask_rcnn_r50_fpn_gn_2x/model_final

configs/gn/faster_rcnn_r50_fpn_gn_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/gn/mask_rcnn_r50_fpn_gn_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/mask_rcnn_r50_fpn_gn_2x/model_final

configs/gridmask/faster_rcnn_r50_vd_fpn_gridmask_4x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 360000
 snapshot_iter: 40000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_gridmask_4x/model_final

configs/hrnet/faster_rcnn_hrnetv2p_w18_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/HRNet_W18_C_pretrained.tar
 weights: output/faster_rcnn_hrnetv2p_w18_1x/model_final

configs/hrnet/faster_rcnn_hrnetv2p_w18_2x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/HRNet_W18_C_pretrained.tar
 weights: output/faster_rcnn_hrnetv2p_w18_2x/model_final

configs/htc/htc_r50_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: HybridTaskCascade
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 50
+log_iter: 50
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/iou_loss/faster_rcnn_r50_vd_fpn_ciou_loss_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_diou_loss_1x/model_final

configs/iou_loss/faster_rcnn_r50_vd_fpn_diou_loss_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_diou_loss_1x/model_final

configs/iou_loss/faster_rcnn_r50_vd_fpn_giou_loss_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_giou_loss_1x/model_final

configs/libra_rcnn/libra_rcnn_r101_vd_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/libra_rcnn_r101_vd_fpn_1x/model_final

configs/libra_rcnn/libra_rcnn_r50_vd_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/libra_rcnn_r50_vd_fpn_1x/model_final

configs/mask_rcnn_r101_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.tar
 metric: COCO

configs/mask_rcnn_r101_vd_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_pretrained.tar
 weights: output/mask_rcnn_r101_vd_fpn_1x/model_final

configs/mask_rcnn_r50_1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/mask_rcnn_r50_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 360000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/mask_rcnn_r50_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 metric: COCO

configs/mask_rcnn_r50_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/mask_rcnn_r50_fpn_2x/model_final

configs/mask_rcnn_r50_vd_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 use_gpu: true
 max_iters: 360000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 metric: COCO

configs/mask_rcnn_se154_vd_fpn_s1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 260000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/SENet154_vd_pretrained.tar
 weights: output/mask_rcnn_se154_vd_fpn_s1x/model_final

configs/mask_rcnn_x101_vd_64x4d_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 180000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar
 weights: output/mask_rcnn_x101_vd_64x4d_fpn_1x/model_final

configs/mask_rcnn_x101_vd_64x4d_fpn_2x.yml

@@ -2,7 +2,7 @@ architecture: MaskRCNN
 max_iters: 360000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar
 weights: output/mask_rcnn_x101_vd_64x4d_fpn_2x/model_final

configs/mobile/cascade_rcnn_mobilenetv3_fpn_320.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNN
 max_iters: 500000
 snapshot_iter: 50000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x1_0_ssld_pretrained.tar
 weights: output/cascade_rcnn_mobilenetv3_fpn_320/model_final

configs/mobile/cascade_rcnn_mobilenetv3_fpn_640.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNN
 max_iters: 500000
 snapshot_iter: 50000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x1_0_ssld_pretrained.tar
 weights: output/cascade_rcnn_mobilenetv3_fpn_640/model_final

configs/obj365/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNNClsAware
 max_iters: 800000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet200_vd_pretrained.tar
 weights: output/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms/model_final

configs/obj365/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas.yml

@@ -3,7 +3,6 @@ max_iters: 500000
 snapshot_iter: 10000
 use_gpu: true
 log_iter: 20
-log_smooth_window: 20
 save_dir: output
 pretrain_weights: https://paddlemodels.bj.bcebos.com/object_detection/cascade_mask_rcnn_dcnv2_se154_vd_fpn_gn_coco_pretrained.tar
 weights: output/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas/model_final

configs/oidv5/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml

@@ -2,7 +2,7 @@ architecture: CascadeRCNNClsAware
 max_iters: 1500000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet200_vd_pretrained.tar
 weights: output/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms/model_final

configs/ppyolo/ppyolo.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 250000
-log_smooth_window: 100
 log_iter: 100
 save_dir: output
 snapshot_iter: 10000

configs/ppyolo/ppyolo_2x.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 100
 log_iter: 100
 save_dir: output
 snapshot_iter: 10000

configs/ppyolo/ppyolo_2x_bs12.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 100
 log_iter: 100
 save_dir: output
 snapshot_iter: 10000

configs/ppyolo/ppyolo_r18vd.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 250000
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 snapshot_iter: 10000

configs/ppyolo/ppyolo_test.yml

@@ -3,7 +3,6 @@
 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 100
 log_iter: 100
 save_dir: output
 snapshot_iter: 10000

configs/random_erasing/faster_rcnn_r50_vd_fpn_random_erasing_4x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 360000
 snapshot_iter: 40000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar
 weights: output/faster_rcnn_r50_vd_fpn_random_erasing_4x/model_final

configs/rcnn_enhance/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 270000
 snapshot_iter: 30000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_ssld_pretrained.tar

configs/rcnn_enhance/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 270000
 snapshot_iter: 30000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar

configs/rcnn_enhance/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.yml

@@ -2,7 +2,6 @@ architecture: FasterRCNN
 max_iters: 270000
 snapshot_iter: 30000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar

configs/rcnn_enhance/generic/cascade_rcnn_cbr101_vd_fpn_server_side.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 1500000
 snapshot_iter: 100000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/CBResNet101_vd_ssld_pretrained.tar

configs/rcnn_enhance/generic/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 1500000
 snapshot_iter: 100000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_vd_ssld_pretrained.tar

configs/rcnn_enhance/generic/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.yml

@@ -2,7 +2,6 @@ architecture: CascadeRCNN
 max_iters: 750000
 snapshot_iter: 50000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_v2_pretrained.tar

configs/res2net/faster_rcnn_res2net50_vb_26w_4s_fpn_1x.yml

@@ -2,7 +2,7 @@ architecture: FasterRCNN
 max_iters: 90000
 snapshot_iter: 10000
 use_gpu: true
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/Res2Net50_26w_4s_pretrained.tar
 weights: output/faster_rcnn_res2net50_vb_26w_4s_fpn_1x/model_final

configs/res2net/mask_rcnn_res2net50_vb_26w_4s_fpn_2x.yml

@@ -3,7 +3,7 @@ trarchitecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/Res2Net50_26w_4s_pretrained.tar
 metric: COCO

configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_2x.yml

@@ -3,7 +3,7 @@ trarchitecture: MaskRCNN
 use_gpu: true
 max_iters: 180000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/Res2Net50_vd_26w_4s_pretrained.tar
 metric: COCO

configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_dcnv2_1x.yml

@@ -3,7 +3,7 @@ trarchitecture: MaskRCNN
 use_gpu: true
 max_iters: 90000
 snapshot_iter: 10000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/Res2Net50_vd_26w_4s_pretrained.tar
 metric: COCO

configs/retinanet_r101_fpn_1x.yml

@@ -3,7 +3,7 @@ max_iters: 90000
 use_gpu: true
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.tar
 weights: output/retinanet_r101_fpn_1x/model_final
-log_smooth_window: 20
+log_iter: 20
 snapshot_iter: 10000
 metric: COCO
 save_dir: output

configs/retinanet_r50_fpn_1x.yml

@@ -3,7 +3,7 @@ max_iters: 90000
 use_gpu: true
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar
 weights: output/retinanet_r50_fpn_1x/model_final
-log_smooth_window: 20
+log_iter: 20
 snapshot_iter: 10000
 metric: COCO
 save_dir: output

configs/retinanet_x101_vd_64x4d_fpn_1x.yml

@@ -3,7 +3,6 @@ max_iters: 180000
 use_gpu: true
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/ResNeXt101_vd_64x4d_pretrained.tar
 weights: output/retinanet_x101_vd_64x4d_fpn_1x/model_final
-log_smooth_window: 20
 log_iter: 20
 snapshot_iter: 30000
 metric: COCO

configs/ssd/ssd_mobilenet_v1_voc.yml

@@ -3,7 +3,7 @@ pretrain_weights: https://paddlemodels.bj.bcebos.com/object_detection/ssd_mobile
 use_gpu: true
 max_iters: 28000
 snapshot_iter: 2000
-log_smooth_window: 1
+log_iter: 1
 metric: VOC
 map_type: 11point
 save_dir: output

configs/ssd/ssd_vgg16_300.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/VGG16_caffe_pretrained.tar

configs/ssd/ssd_vgg16_300_voc.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 120001
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: VOC
 map_type: 11point

configs/ssd/ssd_vgg16_512.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/VGG16_caffe_pretrained.tar

configs/ssd/ssd_vgg16_512_voc.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 120000
 snapshot_iter: 10000
-log_smooth_window: 20
 log_iter: 20
 metric: VOC
 map_type: 11point

configs/ssd/ssdlite_ghostnet.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/GhostNet_x1_3_ssld_pretrained.tar

configs/ssd/ssdlite_mobilenet_v1.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV1_ssld_pretrained.tar

configs/ssd/ssdlite_mobilenet_v3_large.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x1_0_ssld_pretrained.tar

configs/ssd/ssdlite_mobilenet_v3_large_fpn.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x1_0_ssld_pretrained.tar

configs/ssd/ssdlite_mobilenet_v3_small.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_small_x1_0_ssld_pretrained.tar

configs/ssd/ssdlite_mobilenet_v3_small_fpn.yml

@@ -2,7 +2,6 @@ architecture: SSD
 use_gpu: true
 max_iters: 400000
 snapshot_iter: 20000
-log_smooth_window: 20
 log_iter: 20
 metric: COCO
 pretrain_weights: https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_small_x1_0_ssld_pretrained.tar

configs/yolov3_darknet.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov3_darknet_voc.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 70000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 2000
 metric: VOC

configs/yolov3_darknet_voc_diouloss.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 70000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 2000
 metric: VOC

configs/yolov3_mobilenet_v1.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov3_mobilenet_v1_fruit.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 20000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 200
 metric: VOC

configs/yolov3_mobilenet_v1_roadsign.yml

@@ -6,8 +6,8 @@ architecture: YOLOv3
 use_gpu: true
   # ### max_iters为最大迭代次数，而一个iter会运行batch_size * device_num张图片。batch_size在下面 TrainReader.batch_size设置。
 max_iters: 1200
-# log平滑参数，平滑窗口大小，会从取历史窗口中取log_smooth_window大小的loss求平均值
-log_smooth_window: 20
+# log平滑参数，平滑窗口大小，会从取历史窗口中取log_iter大小的loss求平均值
+log_iter: 20
 # 模型保存文件夹
 save_dir: output
 # 每隔多少迭代保存模型

configs/yolov3_mobilenet_v1_voc.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 70000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 2000
 metric: VOC

configs/yolov3_mobilenet_v3.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov3_r34.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 500000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov3_r34_voc.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 70000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 2000
 metric: VOC

configs/yolov4/yolov4_cspdarknet.yml

 architecture: YOLOv4
 use_gpu: true
 max_iters: 500200
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov4/yolov4_cspdarknet_coco.yml

 architecture: YOLOv4
 use_gpu: true
 max_iters: 500200
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 10000
 metric: COCO

configs/yolov4/yolov4_cspdarknet_voc.yml

 architecture: YOLOv4
 use_gpu: true
 max_iters: 140000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 1000
 metric: VOC

contrib/PedestrianDetection/pedestrian_yolov3_darknet.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 200000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 5000
 metric: COCO

contrib/VehicleDetection/vehicle_yolov3_darknet.yml

 architecture: YOLOv3
 use_gpu: true
 max_iters: 120000
-log_smooth_window: 20
+log_iter: 20
 save_dir: output
 snapshot_iter: 2000
 metric: COCO

docs/advanced_tutorials/MODEL_TECHNICAL.md

@@ -290,8 +290,7 @@ OptimizerBuilder:
 ```yaml
 use_gpu: true         # 是否使用GPU运行程序
 max_iters: 500200     # 最大迭代轮数
-log_smooth_window: 20 # 日志打印队列长度
-log_iter: 20          # 训练时日志每迭代x轮打印一次
+log_iter: 20          # 日志打印队列长度, 训练时日志每迭代x轮打印一次
 save_dir: output      # 模型保存路径
 snapshot_iter: 2000   # 训练时第x轮保存/评估
 metric: COCO          # 数据集名称

docs/advanced_tutorials/config_doc/RCNN_PARAMS_DOC.md

@@ -16,9 +16,7 @@ use_gpu: true
 max_iters: 180000
 # 模型保存间隔，如果训练时eval设置为True，会在保存后进行验证
 snapshot_iter: 10000
-# 输出指定区间的平均结果，默认20，即输出20次的平均结果
-log_smooth_window: 20
-# 默认打印log的间隔，默认20
+# 输出指定区间的平均结果，默认20，即输出20次的平均结果。也是默认打印log的间隔。
 log_iter: 20
 # 训练权重的保存路径
 save_dir: output

docs/advanced_tutorials/config_doc/yolov3_mobilenet_v1.md

@@ -8,8 +8,8 @@ architecture: YOLOv3
 use_gpu: true
   # ### max_iters为最大迭代次数，而一个iter会运行batch_size * device_num张图片。batch_size在下面 TrainReader.batch_size设置。
 max_iters: 1200
-# log平滑参数，平滑窗口大小，会从取历史窗口中取log_smooth_window大小的loss求平均值
-log_smooth_window: 20
+# log平滑参数，平滑窗口大小，会从取历史窗口中取log_iter大小的loss求平均值，也是默认打印日志的间隔。
+log_iter: 20
 # 模型保存文件夹
 save_dir: output
 # 每隔多少迭代保存模型

slim/nas/blazeface.yml

 architecture: BlazeFace
 max_iters: 5000
 use_gpu: true
-log_smooth_window: 20
 log_iter: 20
 metric: WIDERFACE
 save_dir: nas_checkpoint

slim/nas/train_nas.py

@@ -328,13 +328,13 @@ def main():
         # When iterable mode, set set_sample_list_generator(train_reader, place)
         train_loader.set_sample_list_generator(train_reader)
 
-        train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
+        train_stats = TrainingStats(cfg.log_iter, train_keys)
         train_loader.start()
         end_time = time.time()
 
         cfg_name = os.path.basename(FLAGS.config).split('.')[0]
         save_dir = os.path.join(cfg.save_dir, cfg_name)
-        time_stat = deque(maxlen=cfg.log_smooth_window)
+        time_stat = deque(maxlen=cfg.log_iter)
         ap = 0
         for it in range(start_iter, cfg.max_iters):
             start_time = end_time

slim/prune/prune.py

@@ -235,14 +235,14 @@ def main():
     # if map_type not set, use default 11point, only use in VOC eval
     map_type = cfg.map_type if 'map_type' in cfg else '11point'
 
-    train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
+    train_stats = TrainingStats(cfg.log_iter, train_keys)
     train_loader.start()
     start_time = time.time()
     end_time = time.time()
 
     cfg_name = os.path.basename(FLAGS.config).split('.')[0]
     save_dir = os.path.join(cfg.save_dir, cfg_name)
-    time_stat = deque(maxlen=cfg.log_smooth_window)
+    time_stat = deque(maxlen=cfg.log_iter)
     best_box_ap_list = [0.0, 0]  #[map, iter]
 
     # use VisualDL to log data

slim/quantization/train.py

@@ -223,14 +223,14 @@ def main():
     # if map_type not set, use default 11point, only use in VOC eval
     map_type = cfg.map_type if 'map_type' in cfg else '11point'
 
-    train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
+    train_stats = TrainingStats(cfg.log_iter, train_keys)
     train_loader.start()
     start_time = time.time()
     end_time = time.time()
 
     cfg_name = os.path.basename(FLAGS.config).split('.')[0]
     save_dir = os.path.join(cfg.save_dir, cfg_name)
-    time_stat = deque(maxlen=cfg.log_smooth_window)
+    time_stat = deque(maxlen=cfg.log_iter)
     best_box_ap_list = [0.0, 0]  #[map, iter]
 
     for it in range(start_iter, cfg.max_iters):

tools/configure.py

@@ -55,7 +55,6 @@ MISC_CONFIG = {
     "weights": "<value>",
     "metric": "<value>",
     "map_type": "11point",
-    "log_smooth_window": 20,
     "snapshot_iter": 10000,
     "log_iter": 20,
     "use_gpu": True,

tools/train.py

@@ -219,14 +219,14 @@ def main():
     # if map_type not set, use default 11point, only use in VOC eval
     map_type = cfg.map_type if 'map_type' in cfg else '11point'
 
-    train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
+    train_stats = TrainingStats(cfg.log_iter, train_keys)
     train_loader.start()
     start_time = time.time()
     end_time = time.time()
 
     cfg_name = os.path.basename(FLAGS.config).split('.')[0]
     save_dir = os.path.join(cfg.save_dir, cfg_name)
-    time_stat = deque(maxlen=cfg.log_smooth_window)
+    time_stat = deque(maxlen=cfg.log_iter)
     best_box_ap_list = [0.0, 0]  #[map, iter]
 
     # use VisualDL to log data

tools/train_multi_machine.py

@@ -213,14 +213,14 @@ def main():
     # if map_type not set, use default 11point, only use in VOC eval
     map_type = cfg.map_type if 'map_type' in cfg else '11point'
 
-    train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
+    train_stats = TrainingStats(cfg.log_iter, train_keys)
     train_loader.start()
     start_time = time.time()
     end_time = time.time()
 
     cfg_name = os.path.basename(FLAGS.config).split('.')[0]
     save_dir = os.path.join(cfg.save_dir, cfg_name)
-    time_stat = deque(maxlen=cfg.log_smooth_window)
+    time_stat = deque(maxlen=cfg.log_iter)
     best_box_ap_list = [0.0, 0]  #[map, iter]
 
     # use VisualDL to log data