ppogg/YOLOv5-Lite issues and pull requests

#112 - add autoanchor.py

Pull Request - State: closed - Opened by Alexsdfdfs over 2 years ago

#111 - Update common.py

Pull Request - State: closed - Opened by Alex866850 over 2 years ago

#110 - update cpp demo

Pull Request - State: closed - Opened by ChaucerG over 2 years ago

#109 - Update yolo.py

Pull Request - State: closed - Opened by ChaucerG over 2 years ago

#108 - YOLOv5-Lite do not recognize torch

Issue - State: closed - Opened by EduardoRonchi over 2 years ago - 3 comments

#107 - predictions not coming for custom dataset with quantized yolov5-lite model

Issue - State: closed - Opened by shivadvg19 over 2 years ago - 2 comments

#106 - efficientnet-lite0 model bug

Issue - State: closed - Opened by brooks0519 over 2 years ago - 2 comments

#105 - 您好，1x1 conv支路，并不会为网络带来提升，反而下降了

Issue - State: closed - Opened by beauty-snowman over 2 years ago - 3 comments

#104 - Error after convert onxx

Issue - State: closed - Opened by Houangnt over 2 years ago - 2 comments

#103 - coco数据集上测试检测不到目标是什么原因呢

Issue - State: closed - Opened by raychiu0202 over 2 years ago - 6 comments

#102 - MNN performance much less than NCNN

Issue - State: closed - Opened by bl6g6 over 2 years ago - 7 comments

#101 - PicoDet用coco训练集train报错

Issue - State: closed - Opened by Pillar1989 over 2 years ago - 19 comments
Labels: bug

#100 - speed problem

Issue - State: closed - Opened by Hiwyl over 2 years ago - 10 comments

#99 - fp16测试正常，量化成int8后没有预测框？

Issue - State: closed - Opened by HuKai97 over 2 years ago - 5 comments

#98 - Win10 下 Pytorch->onnx->ncnn后推理结果和Pytorch推理结果相差很大？

Issue - State: closed - Opened by HuKai97 over 2 years ago - 3 comments

#97 - Unsupported slice axes !

Issue - State: closed - Opened by zqin1998 over 2 years ago - 3 comments

#96 - MNN v5-lite-s drive link not working

Issue - State: closed - Opened by vk1996 over 2 years ago - 7 comments

#95 - Picodet关于neck最后一层的问题

Issue - State: closed - Opened by Alexsdfdfs over 2 years ago - 2 comments

#94 - How to convert custom trained .pt file to ncnn/mnn?

Issue - State: closed - Opened by foemre over 2 years ago - 1 comment

#93 - 关于model hub中的模型和预训练权重的问题

Issue - State: closed - Opened by Lhj-cd over 2 years ago - 2 comments

#92 - unexpected result

Issue - State: closed - Opened by natxopedreira over 2 years ago - 6 comments

#91 - Picodet models have some problems

Issue - State: closed - Opened by triple-Mu over 2 years ago - 1 comment
Labels: bug

#90 - repvggblock相关预训练问题？

Issue - State: closed - Opened by HuKai97 over 2 years ago - 4 comments

#89 - v5Lite.yaml训练自己数据集ncnn推理与pytorch推理结果不一致

Issue - State: closed - Opened by Jordanyang over 2 years ago - 3 comments
Labels: bug

#88 - FPS a little high!

Issue - State: closed - Opened by dsantiago over 2 years ago - 2 comments

#87 - Performance comparison with yolov5n

Issue - State: closed - Opened by thuangb over 2 years ago - 4 comments

#86 - Convert onnx model to IR (xml file)

Issue - State: closed - Opened by AlexisPapaioannou over 2 years ago - 2 comments

#85 - Add PicoDet

Issue - State: closed - Opened by ppogg almost 3 years ago - 8 comments
Labels: enhancement

#84 - 训练存在的问题

Issue - State: closed - Opened by prio158 almost 3 years ago - 10 comments

#83 - pytorch to onnx error

Issue - State: closed - Opened by gengxqa almost 3 years ago - 2 comments

#82 - some errors in running detect.py

Issue - State: closed - Opened by aybberrada almost 3 years ago - 11 comments

#81 - Update README.md (Update broken link)

Pull Request - State: closed - Opened by dankernel almost 3 years ago

#80 - openvion的C++推理

Issue - State: closed - Opened by OpenCDR almost 3 years ago - 1 comment

#79 - 导出onnx 错误

Issue - State: closed - Opened by ChungTak almost 3 years ago - 2 comments

#78 - update ncnn c++ code

Pull Request - State: closed - Opened by dankernel almost 3 years ago

#77 - ‎

Issue - State: closed - Opened by ghost almost 3 years ago - 1 comment

#76 - Update ncnn/README.md

Pull Request - State: closed - Opened by dankernel almost 3 years ago

#75 - v5lite-s-int8.bin file does not exist.

Issue - State: closed - Opened by dankernel almost 3 years ago - 2 comments

#74 - Update README.md

Pull Request - State: closed - Opened by dankernel almost 3 years ago

#73 - 您好，model hub超链接怎么失效了呢？

Issue - State: closed - Opened by Wanghe1997 almost 3 years ago - 2 comments

#72 - Smaller model?

Issue - State: closed - Opened by Msyout almost 3 years ago

#71 - 手机移动端详细教程

Issue - State: closed - Opened by Alexsdfdfs almost 3 years ago

#70 - 树莓派如何部署ncnn

Issue - State: closed - Opened by Liaruto almost 3 years ago - 2 comments

#69 - openvino inference error

Issue - State: closed - Opened by Galaxy-yyyyy almost 3 years ago - 2 comments

#68 - Output of v5Lite-s model

Issue - State: closed - Opened by AlexisPapaioannou almost 3 years ago - 4 comments

#67 - Trying v5Lite-s.pt on a single image

Issue - State: closed - Opened by kennybradley almost 3 years ago - 2 comments

#66 - Have you tried pruning YOLOv5-Liteg through sparsity training and channel pruning

Issue - State: closed - Opened by HouBiaoLiu almost 3 years ago - 1 comment

#65 - 模型文件在master版本的yolo下面跑不起来

Issue - State: closed - Opened by deep-practice almost 3 years ago - 7 comments

#64 - 转换脚本 scripts/rep_convert.py 重参化g模型的存在问题

Issue - State: closed - Opened by Xiaoyw1998 almost 3 years ago - 3 comments

#63 - 在COCO上直接训练v5lite-s 416x416，未修改任何参数，map仅35.2

Issue - State: closed - Opened by Broad-sky almost 3 years ago - 12 comments
Labels: documentation

#62 - 没有找到weights文件夹

Issue - State: closed - Opened by liufan19960306 almost 3 years ago - 4 comments

#61 - voc 上的精度有数据吗

Issue - State: closed - Opened by yokings almost 3 years ago - 5 comments

#60 - Got Segmentation Fault when running ncnn on Raspberry Pi

Issue - State: closed - Opened by CWilliamL almost 3 years ago - 5 comments

#59 - 导出onnx并用opencv dnn调用满屏都是框

Issue - State: closed - Opened by planeflysun almost 3 years ago - 6 comments

#58 - 复现不出RepVGG(Silu)的精度

Issue - State: closed - Opened by sunriselqq almost 3 years ago - 13 comments

#57 - 转tensorrt 报错

Issue - State: open - Opened by woshituobaye almost 3 years ago - 19 comments

#56 - ShuffleBlock Convert to Tensorflow model

Issue - State: closed - Opened by AlexisPapaioannou almost 3 years ago - 2 comments

#55 - onnx输出不包括detect层中的inference内容，也没有将输出合并

Issue - State: closed - Opened by cxnaive almost 3 years ago - 3 comments

#54 - 使用openvino推理xml文件报错

Issue - State: closed - Opened by ghost almost 3 years ago - 8 comments

#53 - ONNX导出NCNN模型的问题解决+完整int8量化步骤

Issue - State: closed - Opened by ChaucerG almost 3 years ago - 24 comments
Labels: good first issue

#52 - 报错dataset not found

Issue - State: closed - Opened by suhanling almost 3 years ago - 9 comments

#51 - 找不到权重文件

Issue - State: closed - Opened by suhanling almost 3 years ago - 2 comments

#50 - 关于openvino量化的问题

Issue - State: closed - Opened by cxnaive almost 3 years ago - 1 comment

#49 - YOLOv5-Lite v1.4

Issue - State: closed - Opened by ppogg almost 3 years ago
Labels: enhancement

#48 - YOLOv5-Lite v1.4

Issue - State: closed - Opened by ppogg almost 3 years ago
Labels: enhancement

#47 - 怎么没有v5Lite-c.pt的预训练模型？

Issue - State: closed - Opened by youbinqiao almost 3 years ago - 4 comments

#46 - weights folder no long supplied

Issue - State: closed - Opened by prying almost 3 years ago - 3 comments

#45 - 预存连模型链接挂了

Issue - State: closed - Opened by cxnaive almost 3 years ago - 4 comments

#44 - 关于c++部署的问题？

Issue - State: closed - Opened by Hezhexi2002 almost 3 years ago

#43 - 关于使用c++部署的问题？

Issue - State: closed - Opened by Hezhexi2002 almost 3 years ago - 10 comments

#42 - mnn demo error

Issue - State: closed - Opened by abcliguanxi almost 3 years ago - 2 comments

#41 - 您好，请问如何做VOC数据集上的训练工作？

Issue - State: closed - Opened by fantow almost 3 years ago - 1 comment

#38 - the download links of v5lite-s.pt and v5lite-g.pt are invalid

Issue - State: closed - Opened by zhaojc001 almost 3 years ago - 1 comment

#32 - Light_C3 function

Issue - State: closed - Opened by AlexisPapaioannou almost 3 years ago - 4 comments

#31 - What happen with my yolov5.param file ?

Issue - State: closed - Opened by Houangnt almost 3 years ago - 10 comments

#30 - where is "layer.h"

Issue - State: closed - Opened by aimen123 almost 3 years ago - 3 comments

#29 - ncnn raspberry pi example

Issue - State: closed - Opened by JNaranjo-Alcazar almost 3 years ago - 1 comment

#28 - Problem converting yolov5-lite to tflite

Issue - State: closed - Opened by JNaranjo-Alcazar almost 3 years ago - 4 comments

#27 - Which model is the fastest?

Issue - State: closed - Opened by JNaranjo-Alcazar about 3 years ago - 1 comment

#26 - Tensorflow Converter

Issue - State: closed - Opened by AlexisPapaioannou about 3 years ago - 5 comments

#25 - 没有提供coco权重的.pt文件吗？

Issue - State: closed - Opened by violet2422 about 3 years ago - 4 comments

#24 - ncnn测试很慢的问题

Issue - State: closed - Opened by chenweifu2008 about 3 years ago - 6 comments

#23 - 你好，请问这个代码检测目标置信度的阈值在什么地方可以设置？

Issue - State: closed - Opened by Peppachou about 3 years ago - 1 comment

#22 - the speed is not fast?

Issue - State: closed - Opened by henbucuoshanghai about 3 years ago - 21 comments

#21 - 你好请问训练细节和yolo5一样？只需要yaml文件改对就可以？核心后处理一样吗？

Issue - State: closed - Opened by henbucuoshanghai about 3 years ago - 3 comments

#20 - 是否考虑p6的模型

Issue - State: closed - Opened by yongjingli about 3 years ago - 12 comments

#17 - 未提供yolov5ss-0.5的速度和map

Issue - State: closed - Opened by Msyout about 3 years ago - 1 comment

#13 - Here is a performance table that evaluate by coco api

Issue - State: closed - Opened by ppogg about 3 years ago
Labels: enhancement

#12 - 批量预测保存

Issue - State: closed - Opened by ppflyba-zm about 3 years ago - 1 comment

#11 - How to save the r, p, f1 of plot_mc_curve in txt

Issue - State: closed - Opened by larrywal-express about 3 years ago - 5 comments

#10 - 测试帧率问题

Issue - State: closed - Opened by luckywangchenxi about 3 years ago - 7 comments

#9 - Create LICENSE

Pull Request - State: closed - Opened by ppogg about 3 years ago

#8 - 训不到大佬的结果

Issue - State: closed - Opened by hyaihjq about 3 years ago - 2 comments

#7 - 请问能否将模型发布在torch hub上？

Issue - State: closed - Opened by rener2020 about 3 years ago - 3 comments

#6 - 为什么不用mobilenetv3

Issue - State: closed - Opened by Msyout about 3 years ago - 3 comments

#5 - 多卡分布式训练缓慢

Issue - State: closed - Opened by Songrtsu about 3 years ago - 1 comment

#4 - 想知道这个模型对少样本数据如何?

Issue - State: closed - Opened by Alex866850 about 3 years ago - 3 comments

#3 - shufflenetv2预训练的模型的问题

Issue - State: closed - Opened by chenweifu2008 about 3 years ago - 7 comments

#2 - 大佬能提供训练的yolov4-tiny、yolov3-tiny的链接或者代码吗？

Issue - State: closed - Opened by ahhhxxx about 3 years ago - 5 comments

#1 - 关于ONNX调用的问题

Issue - State: closed - Opened by Olalaye about 3 years ago - 1 comment

GitHub / ppogg/YOLOv5-Lite issues and pull requests