Skip to content

Issues: Tencent/ncnn

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

1,066 Open 2,503 Closed
1,066 Open 2,503 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

SoftMax3D enhancement
#5823 opened Dec 15, 2024 by DickyQi
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 Dec 13, 2024 by hjq052070
1
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 Dec 13, 2024 by cslvjt
3
nanodet plus 1152*768 转换成功，结果不匹配
#5819 opened Dec 11, 2024 by Yindong-Zhang
2
Compatibility with RT-DETR
#5817 opened Dec 9, 2024 by shery909
HarmonyOS next系统ncnn无法识别图片
#5816 opened Dec 9, 2024 by WylAndy
pnnx转换后ncnn.py不能正常推理运行
#5813 opened Dec 4, 2024 by HongFaWan
有一些算子不支持
#5812 opened Dec 3, 2024 by HongFaWan
pnnx转换模型生成的ncnn模型shape不对，pnnx模型shape正确
#5810 opened Dec 3, 2024 by terrencew
layer ReduceL2 not exists or registered
#5808 opened Nov 29, 2024 by YYY7C
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 Nov 28, 2024 by emptyery
3
不支持index_select算子
#5802 opened Nov 27, 2024 by qingyang-jiandao
编译的hmos版本的静态库在使用时报错neither ET_REL nor LLVM bitcode
#5801 opened Nov 27, 2024 by ahaha4321
pnnx.py and ncnn.py for resnet18 has a slight different final outputs
#5800 opened Nov 26, 2024 by Bing1002
float16 溢出问题
#5799 opened Nov 26, 2024 by qingyang-jiandao
2
谁有yolov8系列的Classification，segment，pose，obb。
#5798 opened Nov 25, 2024 by mingtianquan
Version differences result in different running outcomes
#5797 opened Nov 25, 2024 by HongFaWan
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 Nov 25, 2024 by tianqi0072600
2
Successfuly run with pnnx, but failed with nccn inference.
#5792 opened Nov 25, 2024 by zengjie617789
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 Nov 21, 2024 by wyb222333
2
onnx转换出错
#5783 opened Nov 20, 2024 by q2547978337
python setup failing
#5782 opened Nov 19, 2024 by xalteropsx
use reID on android
#5780 opened Nov 18, 2024 by Edward-YS
求助：yolo11直接导出的ncnn模型，example代码推理乱框
#5775 opened Nov 13, 2024 by dixiatielu
1
android fp16 use_fp16_arithmetic=true精度下降严重
#5771 opened Nov 8, 2024 by cbingdu
1
ProTip! Adding no:label will show everything without a label.