Tencent/ncnn

Issues

• SoftMax3D

  #5823 opened 7 days ago by DickyQi
  0

• 使用官方的权重和速度评测工具在X86上面，不同模型量化前后的速度提升很大

  #5820 opened 9 days ago by cslvjt
  3

• 模型处理y通道时，图像内容缺失

  #5824 opened 6 days ago by cslvjt
  2

• when I use ncnn model to infer some images, for each image, I must create a new extractor? 每次有新输入都必须创建一个新的extractor吗？

  #5822 opened 9 days ago by hjq052070
  1

• HarmonyOS next系统ncnn无法识别图片

  #5816 opened 13 days ago by WylAndy
  0

• nanodet plus 1152*768 转换成功，结果不匹配

  #5819 opened 11 days ago by Yindong-Zhang
  2

• 使用量化后的模型，app闪退

  #5818 opened 11 days ago by CangHaiQingYue
  0

• pnnx转换模型生成的ncnn模型shape不对，pnnx模型shape正确

  #5810 opened 19 days ago by terrencew
  0

• Compatibility with RT-DETR

  #5817 opened 13 days ago by shery909
  0

• Resnet50 not working

  #5790 opened 14 days ago by xalteropsx
  0

• yolov5转.torchscript通过PNNX转ncnn，华为，小米手机输出结果正常，一加AcePro，识别错误。但是一加采用同模型的onnx推理结果正确的

  #5785 opened a month ago by wyb222333
  2

• pnnx转换后ncnn.py不能正常推理运行

  #5813 opened 18 days ago by HongFaWan
  0

• 有一些算子不支持

  #5812 opened 19 days ago by HongFaWan
  0

• rtdetr-l转ncnn

  #5765 opened 2 months ago by x825196964
  1

• The NCDHW input required for the 3DCNN model is modified to DCHW in NCNN, right?

  #5809 opened 23 days ago by erquren
  1

• 新手求解，用pnnx把yolov10从torchscript模型转到ncnn模型时算子不支持

  #5803 opened 24 days ago by emptyery
  3

• layer ReduceL2 not exists or registered

  #5808 opened 23 days ago by YYY7C
  0

• 20240820版本本地运行崩溃

  #5776 opened a month ago by leizh007
  5

• 不支持index_select算子

  #5802 opened 25 days ago by qingyang-jiandao
  0

• 编译的hmos版本的静态库在使用时报错neither ET_REL nor LLVM bitcode

  #5801 opened 25 days ago by ahaha4321
  0

• pnnx.py and ncnn.py for resnet18 has a slight different final outputs

  #5800 opened a month ago by Bing1002
  0

• 自己训练的模型yolo11导出ncnn格式文件后，部署在安卓上闪退；而官方自带导出后能正常运行，可能是什么问题呢？

  #5795 opened a month ago by tianqi0072600
  2

• float16 溢出问题

  #5799 opened a month ago by qingyang-jiandao
  2

• Successfuly run with pnnx, but failed with nccn inference.

  #5792 opened a month ago by zengjie617789
  0

• 谁有yolov8系列的Classification，segment，pose，obb。

  #5798 opened a month ago by mingtianquan
  0

• Version differences result in different running outcomes

  #5797 opened a month ago by HongFaWan
  0

• The tools directory is not in build

  #5796 opened a month ago by hl-ux
  1

• onnx 的 LayerNormalization 和 ncnn的LayerNorm 有啥区别吗

  #5793 opened a month ago by cucibala
  0

• how long it take to work?

  #5781 opened a month ago by xalteropsx
  0

• onnx转换出错

  #5783 opened a month ago by q2547978337
  0

• python setup failing

  #5782 opened a month ago by xalteropsx
  0

• rgba image cannot be process why is that

  #5756 opened a month ago by xalteropsx
  0

• windows端在Qt Creator中引入ncnn库，运行时报"platform.h"中的mutex错误

  #5777 opened a month ago by zhahoi
  7

• use reID on android

  #5780 opened a month ago by Edward-YS
  0

• need yolo inference code

  #5772 opened a month ago by xalteropsx
  6

• 求助：yolo11直接导出的ncnn模型，example代码推理乱框

  #5775 opened a month ago by dixiatielu
  1

• 求助：从pth转为ncnn模型，推理出来的结果不一样

  #5768 opened a month ago by zengweigit
  2

• 现在onnx模型怎么转化为ncnn模型呢？

  #5770 opened a month ago by HDL199411
  1

• android fp16 use_fp16_arithmetic=true精度下降严重

  #5771 opened a month ago by cbingdu
  1

• nvm my bad

  #5769 opened a month ago by xalteropsx
  0

• 缺少torch.einsum("ijm,imkl->ijkl", v_920, v_486)支持

  #5766 opened 2 months ago by Baiyuetribe
  0

• ncnn build keep failing

  #5761 opened 2 months ago by xalteropsx
  2

• @Rachel-liuqr I have great news 😃! I've recently added official support for Ultralytics YOLOv8 NCNN export ✅ in PR https://github.com/ultralytics/ultralytics/pull/3529 with the help of @nihui which is part of `ultralytics==8.0.129`. NCNN works for all tasks including Detect, Segment, Pose and Classify.

  #5764 opened 2 months ago by xkai66
  0

• examples/yolov8.cpp produce random detections on android system

  #5760 opened 2 months ago by Deephome
  0

• PNNX转换fastspeech2的onnx模型报错

  #5758 opened 2 months ago by qweasdzxcvde
  2

• 安卓，多次运行deeplabv3 分割模型，结果不一致。

  #5757 opened 2 months ago by zhaoyk1986
  4

• [BUG]gfpgan1.4系列里，部分情况下Reduction算子GPU推理溢出

  #5759 opened 2 months ago by Baiyuetribe
  1

• BinaryOp的加法操作占用大量内存

  #5751 opened 2 months ago by 1072946446
  1

• ncnn failiure

  #5753 opened 2 months ago by xalteropsx
  0

• YoloV5-6.0的c++推理问题

  #5752 opened 2 months ago by PengPeng-JunJun
  0