You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

README.md 3.5 kB

2 years ago
123456789101112131415161718192021222324
  1. # Mixed Precision Training
  2. ## Introduction
  3. <!-- [OTHERS] -->
  4. ```latex
  5. @article{micikevicius2017mixed,
  6. title={Mixed precision training},
  7. author={Micikevicius, Paulius and Narang, Sharan and Alben, Jonah and Diamos, Gregory and Elsen, Erich and Garcia, David and Ginsburg, Boris and Houston, Michael and Kuchaiev, Oleksii and Venkatesh, Ganesh and others},
  8. journal={arXiv preprint arXiv:1710.03740},
  9. year={2017}
  10. }
  11. ```
  12. ## Results and Models
  13. | Architecture | Backbone | Style | Conv | Lr schd | Mem (GB) | Inf time (fps) | box AP | mask AP | Config | Download |
  14. |:------------:|:---------:|:-------:|:------------:|:-------:|:--------:|:--------------:|:------:|:-------:|:------:|:--------:|
  15. | Faster R-CNN | R-50 | pytorch | - | 1x | 3.4 | 28.8 | 37.5 | - |[config](https://github.com/open-mmlab/mmdetection/tree/master/configs/fp16/faster_rcnn_r50_fpn_fp16_1x_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/fp16/faster_rcnn_r50_fpn_fp16_1x_coco/faster_rcnn_r50_fpn_fp16_1x_coco_20200204-d4dc1471.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/fp16/faster_rcnn_r50_fpn_fp16_1x_coco/faster_rcnn_r50_fpn_fp16_1x_coco_20200204_143530.log.json) |
  16. | Mask R-CNN | R-50 | pytorch | - | 1x | 3.6 | 24.1 | 38.1 | 34.7 |[config](https://github.com/open-mmlab/mmdetection/tree/master/configs/fp16/mask_rcnn_r50_fpn_fp16_1x_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_1x_coco/mask_rcnn_r50_fpn_fp16_1x_coco_20200205-59faf7e4.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_1x_coco/mask_rcnn_r50_fpn_fp16_1x_coco_20200205_130539.log.json) |
  17. | Mask R-CNN | R-50 | pytorch | dconv(c3-c5) | 1x | 3.0 | | 41.9 | 37.5 |[config](https://github.com/open-mmlab/mmdetection/tree/master/configs/fp16/mask_rcnn_r50_fpn_fp16_dconv_c3-c5_1x_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_dconv_c3-c5_1x_coco/mask_rcnn_r50_fpn_fp16_dconv_c3-c5_1x_coco_20210520_180247-c06429d2.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_dconv_c3-c5_1x_coco/mask_rcnn_r50_fpn_fp16_dconv_c3-c5_1x_coco_20210520_180247.log.json) |
  18. | Mask R-CNN | R-50 | pytorch | mdconv(c3-c5)| 1x | 3.1 | | 42.0 | 37.6 |[config](https://github.com/open-mmlab/mmdetection/tree/master/configs/fp16/mask_rcnn_r50_fpn_fp16_mdconv_c3-c5_1x_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_mdconv_c3-c5_1x_coco/mask_rcnn_r50_fpn_fp16_mdconv_c3-c5_1x_coco_20210520_180434-cf8fefa5.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/fp16/mask_rcnn_r50_fpn_fp16_mdconv_c3-c5_1x_coco/mask_rcnn_r50_fpn_fp16_mdconv_c3-c5_1x_coco_20210520_180434.log.json) |
  19. | Retinanet | R-50 | pytorch | - | 1x | 2.8 | 31.6 | 36.4 | |[config](https://github.com/open-mmlab/mmdetection/tree/master/configs/fp16/retinanet_r50_fpn_fp16_1x_coco.py) | [model](https://download.openmmlab.com/mmdetection/v2.0/fp16/retinanet_r50_fpn_fp16_1x_coco/retinanet_r50_fpn_fp16_1x_coco_20200702-0dbfb212.pth) &#124; [log](https://download.openmmlab.com/mmdetection/v2.0/fp16/retinanet_r50_fpn_fp16_1x_coco/retinanet_r50_fpn_fp16_1x_coco_20200702_020127.log.json) |

No Description

Contributors (3)