-
avg loss always -nan with csresnext50-panet-spp-original-optimal.cfg training after 3000 interations
The training output :
```
Tensor Cores are used. (next mAP calculation at 4200 iterations)
41…
-
hi, I try to use darknet.dll with YOLOV4 after compile darknet,when programme load cfg network,system show me error:"Darknet error location: D:\TRY\darknet-master\src\dark_cuda.c, cudnn_check_error, …
-
![Screenshot from 2021-09-01 12-08-59](https://user-images.githubusercontent.com/49749220/131610369-dbc5e87c-20f9-4d8d-a2d4-531bb0f17ec5.png)
As you can see, some mask is right, but some mask is …
-
I hope this is not a noob question, but i spent the day on trying different things and i just cannot find why darknet training cannot use GPU memory right:
If you want to report a bug - provide:
…
-
then train yolov4,load model success and immediately stop,log:
nms_kind: greedynms (1), beta = 0.600000
151 route 147 -> 26 x 26 x 256
152 conv 512 3 x…
-
Scaled-YOLOv4: Scaling Cross Stage Partial Network - The best neural network for object detection (Top1 accuracy on MS COCO dataset)
Scaled YOLO v4 is the most accurate neural network (55.8% AP Mic…
-
### Prerequisite
- [X] I have searched [the existing and past issues](https://github.com/open-mmlab/mmyolo/issues) but cannot get the expected help.
- [X] I have read the [FAQ documentation](https://…
-
@Zzh-tju thank you for this excellent repo!! **Your comparisons are very well documented and very clear to understand.** I wanted to ask if you had compared Soft-NMS as well? I saw on the efficientdet…
-
After exporting weights using:
./darknet export yolo4tiny.cfg yolo4-tiny_best.weights layers
```
mini_batch = 8, batch = 64, time_steps = 1, train = 1
layer filters size/strd(dil) inp…
-