add 3x_retina

[jshilong]

Motivation

Added 3x multi-scale training config of RetinaNet

retinanet_r50_fpn_mstrain_640-800_3x_coco.py	39.5
retinanet_r101_fpn_mstrain_640-800_3x_coco.py	41
retinanet_caffe_r101_fpn_mstrain_640-800_3x_coco.py	40.7
retinanet_x101_64x4d_fpn_mstrain_640-800_3x_coco.py	41.6

[codecov]

Codecov Report

Merging #5636 (c0c6d6b) into master (0e09f00) will increase coverage by 1.42%.
The diff coverage is n/a.

❗ Current head c0c6d6b differs from pull request most recent head 96d8565. Consider uploading reports for the commit 96d8565 to get more accurate results

@@            Coverage Diff             @@
##           master    #5636      +/-   ##
==========================================
+ Coverage   65.01%   66.44%   +1.42%     
==========================================
  Files         275      281       +6     
  Lines       21262    22019     +757     
  Branches     3537     3659     +122     
==========================================
+ Hits        13824    14631     +807     
+ Misses       6687     6632      -55     
- Partials      751      756       +5     

Flag	Coverage Δ
unittests	66.44% <ø> (+1.46%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmdet/models/detectors/detr.py	45.45% <0.00%> (-54.55%)	⬇️
mmdet/models/backbones/ssd_vgg.py	68.88% <0.00%> (-18.77%)	⬇️
mmdet/models/dense_heads/detr_head.py	74.16% <0.00%> (-15.12%)	⬇️
mmdet/models/utils/normed_predictor.py	83.33% <0.00%> (-9.53%)	⬇️
mmdet/datasets/pipelines/loading.py	50.64% <0.00%> (-7.78%)	⬇️
mmdet/models/detectors/single_stage.py	66.03% <0.00%> (-5.67%)	⬇️
mmdet/models/roi_heads/point_rend_roi_head.py	16.09% <0.00%> (-3.54%)	⬇️
mmdet/models/dense_heads/dense_test_mixins.py	41.02% <0.00%> (-2.57%)	⬇️
mmdet/models/dense_heads/ssd_head.py	90.29% <0.00%> (-2.48%)	⬇️
mmdet/models/backbones/darknet.py	57.69% <0.00%> (-1.29%)	⬇️
... and 61 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0e09f00...96d8565. Read the comment docs.