Ling

https://www.1027663760.com/ [email protected]

能在自己所热爱的平台上做出点东西，挺好

Results 16 comments of


                                            Ling

我有3个GPU，但get_gpu_count()=1

> NCNN的GPU推理是基于vulkan的，可能是你的P40的驱动不支持vulkan导致可以通过vulkaninfo命令查看vulkan识别到的设备 vulkaninfo 也只能看到一个GPU，那这样的话是不是无解了 ================= VK_LAYER_NV_optimus (NVIDIA Optimus layer) Vulkan version 1.3.277, layer version 1: Layer Extensions: count = 0 Devices: count = 1 GPU id = 0 (NVIDIA...

我有3个GPU，但get_gpu_count()=1

> > > NCNN的GPU推理是基于vulkan的，可能是你的P40的驱动不支持vulkan导致可以通过vulkaninfo命令查看vulkan识别到的设备 > > > > > > vulkaninfo 也只能看到一个GPU，那这样的话是不是无解了 > > ================= VK_LAYER_NV_optimus (NVIDIA Optimus layer) Vulkan version 1.3.277, layer version 1: Layer Extensions: count =...

我有3个GPU，但get_gpu_count()=1

> > > > > NCNN的GPU推理是基于vulkan的，可能是你的P40的驱动不支持vulkan导致可以通过vulkaninfo命令查看vulkan识别到的设备 > > > > > > > > > > > > vulkaninfo 也只能看到一个GPU，那这样的话是不是无解了 > > > > ================= VK_LAYER_NV_optimus (NVIDIA Optimus layer)...

我有3个GPU，但get_gpu_count()=1

> Windows下装谷歌提供的grid驱动就能用图形功能了 grid驱动安装后 3060就认不到了...

YOLOv9 is better than any convolution or transformer based object detectors

Have you considered adapting YOLOV9 for Darknet?

CUDNN error at training iteration 1000 when calculating mAP%

The latest version of cudnn always has various bugs

@AlexeyAB Is this project no longer maintained?

> Alexey is no longer on this project. See the FAQ for details: https://www.ccoderun.ca/programming/yolo_faq/#which_repos_to_use_and_avoid https://github.com/hank-ai/darknet Have you considered some big changes to Darknet? For example, support int8

Yolov9

Compared to other frameworks, Darknet does not have too many dependencies and is easy to integrate. I hope to see major updates to Darknet in the future

好兄弟，啥时候有yolov8n/s/m.cfg

> 本周末 yolov8跑不起来呢 ![QQ截图20230319144911](https://user-images.githubusercontent.com/26378634/226158996-78d96dab-2a93-4508-a90f-690ea2c8c9ad.png)

好兄弟，啥时候有yolov8n/s/m.cfg

> Debuged 我试了yolov5和8的cfg，都是训练到一半就 -nan了，yolov4-tiny 一直正常

1
2
›