|
|
@@ -1,21 +1,5 @@
|
|
|
# YOLOv5:
|
|
|
|
|
|
-| Model | Batch | Scale | AP<sup>val<br>0.5:0.95 | AP<sup>val<br>0.5 | FLOPs<br><sup>(G) | Params<br><sup>(M) | Weight |
|
|
|
-|-----------|-------|-------|------------------------|-------------------|-------------------|--------------------|--------|
|
|
|
-| YOLOv5-N | 1xb16 | 640 | 29.8 | 47.1 | 7.7 | 2.4 | [ckpt](https://github.com/yjh0410/RT-ODLab/releases/download/yolo_tutorial_ckpt/yolov5_n_coco.pth) |
|
|
|
-| YOLOv5-S | 1xb16 | 640 | 37.8 | 56.5 | 27.1 | 9.0 | [ckpt](https://github.com/yjh0410/RT-ODLab/releases/download/yolo_tutorial_ckpt/yolov5_s_coco.pth) |
|
|
|
-| YOLOv5-M | 1xb16 | 640 | 43.5 | 62.5 | 74.3 | 25.4 | [ckpt](https://github.com/yjh0410/RT-ODLab/releases/download/yolo_tutorial_ckpt/yolov5_m_coco.pth) |
|
|
|
-| YOLOv5-L | 1xb16 | 640 | 46.7 | 65.5 | 155.6 | 54.2 | [ckpt](https://github.com/yjh0410/RT-ODLab/releases/download/yolo_tutorial_ckpt/yolov5_l_coco.pth) |
|
|
|
-
|
|
|
-- For training, we train YOLOv5 series with 300 epochs on COCO.
|
|
|
-- For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation, following the setting of [YOLOv5](https://github.com/ultralytics/yolov5).
|
|
|
-- For optimizer, we use SGD with weight decay 0.0005 and base per image lr 0.01 / 64, following the setting of the official YOLOv5.
|
|
|
-- For learning rate scheduler, we use linear decay scheduler.
|
|
|
-- We use decoupled head in our reproduced YOLOv5, which is different from the official YOLOv5'head.
|
|
|
-
|
|
|
-
|
|
|
-On the other hand, we are trying to use **AdamW** and larger batch size to train our reproduced YOLOv5. We will update the new results as soon as possible.
|
|
|
-
|
|
|
| Model | Batch | Scale | AP<sup>val<br>0.5:0.95 | AP<sup>val<br>0.5 | FLOPs<br><sup>(G) | Params<br><sup>(M) | Weight |
|
|
|
|-----------|-------|-------|------------------------|-------------------|-------------------|--------------------|--------|
|
|
|
| YOLOv5-N | 8xb16 | 640 | | | | | |
|
yjh0410