学无止境、积少成多、厚积薄发

2024Bug记录

记录2023年遇到的Bug以及修复方案

ValueError: Solver lbfgs supports only ‘l2‘ or ‘none‘ penalties, got l1 penalty

Could not locate zlibwapi.dll. Please make sure it is in your library path!

在3.0以上python中，input()函数，通过键盘输入返回值的类型是字符串，要用int、float等进行强制类型转换。因为，input()函数是默认输入字符串，不管你输入的数字还是其他什么。。。具体例子：Test.pym = input()n =input()res = m*nprint(res)修改：m = input()n =input()res = int(m...

OSError: [WinError 1455] 页面文件太小，无法完成操作。 Error loading "D:\anaconda3\envs\yolov5\lib\site-packages\torch\lib\shm.dll" or one of its dependencies.

无法加载文件 C:\Users\Documents\WindowsPowerShell\profile.ps1

ImportError: cannot import name 'InterruptiblePdb' from partially initialized module 'IPython.core.debugger' (most likely due to a circular import) (D:\anaconda3\envs\data\lib\site-packages\IPython\core\debugger.py)[W 13:17:20.179 NotebookApp] KernelResta

成功解决将XGBoost中plot_importance绘图时出现的f0、f1、f2、f3、f4、f5等改为对应特征的字段名。

Downgrade the protobuf package to 3.20.x or lower.

ERROR: Could not build wheels for opencv-python, which is required to install pyproject.toml-based projects

原因是pip安装的是最新的xlrd包，只支持.xls文件。所以pd.read_excel('xxx.xlsx')会报错。解决方法1: 安装旧版xlrdpip uninstall xlrdpip install xlrd=1.2.0解决方案2: 使用openpyxl包pd.read_excel('xxx.xlsx', engine='openpyxl')...

在本地机器上，高频率重复调用一个API接口，出现“Max retries exceeded with url”，拒绝连接的情况。楼主讲一下，遇到这个bug的过程：在服务器上开发了一个识别图像文字API接口，也部署好了，准备对接口进行测试，楼主准备对本地文件夹中的2000多张图片进行测试，也就是说需要重复调用API接口2000多次。请求接口50多次的时候，接口就不反馈信息了，报错，连接错误，无法连接，一开始以为是网络的问题，然后又重头开始调用直到调了300多次，报错，连接错误，无法连接，一直这么下去大

错误信息：tensor.sub_(mean[:, None, None]).div_(std[:, None, None])RuntimeError: The size of tensor a (4) must match the size of tensor b (3) at non-singleton dimension 0报错原因：维度信息不匹配修改之前：image = Image.open(image_path)修改之后：image = Image.open(image_path)

今天记录一下很伤悲（SB）的bug：AttributeError: module ‘cv2’ has no attribute ‘imread’准备学习一下opencv-python中的resize()函数，然后新建了一个.py脚本import cv2img = cv2.imread("audi.jpg")# 后面还有很多代码，不输入了运行代码，AttributeError: module ‘cv2’ has no attribute ‘imread’此时我一脸懵逼，难道环境崩了？op

在利用下面的语句指定GPU为仅为"1"，但是用nvidia-smi查看GPU使用时，仍默认为"0"import ioimport osimport timeimport jsonimport cv2import numpy as npfrom flask import Flask, jsonify, requestfrom *** import **** # torch等，举个例子os.environ['CUDA_VISIBLE_DEVICES'] = '1'解决方案：将上述语句放

报错如下:密密麻麻的提示，看得我头皮发麻。。。。参考：https://blog.csdn.net/shiwanghualuo/article/details/101756876

import torchimport torchvisionimport torchvision.transforms as transformstransform = transforms.Compose([ transforms.ToTensor(), transforms.Normalize((0.5), (0.5))])trainset = torchvision.datasets.MNIST(root='./data', train=True, download...

peself._apply(transformer, node_wrapper, log_level=index + 1)File “/home/ubuntu/anaconda3/envs/PPYOLO/lib/python3.8/site-packages/paddle/fluid/dygraph/dygraph_to_static/ast_transformer.py”, line 65, in _applytransformer(node_wrapper).transform()File “/

今天部署模型时遇到一个问题：RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same报错原因：原因是cpu和cdua 使用的不一致，我这里是因为模型加载时没用将其分配到GPU中，做个记录。解决办法：device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")prin

有很多xml文件，但是用浏览器打不开，用txt可以打开浏览器显示：记事本显示：报错：xmlParseEntityRef: no name的错误解决方法：最简单的解决方法就是把filename中的&符号去掉

原代码：from PIL import Imageimage = Image.open('steel2.jpg') # 读取image.save('save.jpg') # 保存报错：OSError: cannot write mode RGBA as JPEG原因：RGBA意思是红色，绿色，蓝色，Alpha的色彩空间，Alpha指透明度。而JPG不支持透明度，所以要么丢弃Alpha,要么保存为.png文件解决：方法一：image=image.convert('RGB')image

RuntimeError: main thread is not in main loopException ignored in: <function Image.del at 0x7f5768288940>Traceback (most recent call last):File “/home/ubuntu/anaconda3/envs/torch17/lib/python3.9/tkinter/init.py”, line 4015, in delself.tk.call(‘im

File “/home/myuser/anaconda3/envs/ztf/lib/python3.6/site-packages/keras/engine/topology.py”, line 3328, in load_weights_from_hdf5_grouporiginal_keras_version = f.attrs[‘keras_version’].decode(‘utf8’)AttributeError: ‘str’ object has no attribute ‘decode’.

参考keras中文文档https://keras.io/zh/models/model/https://www.jianshu.com/p/4203a6435ab5https://blog.csdn.net/u012862372/article/details/80367607https://blog.csdn.net/ywcpig/article/details/105659637

qt.qpa.plugin: Could not load the Qt platform plugin “xcb” in “” even though it was found.This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.Available platform plugins a.

Installing build dependencies ... error ERROR: Command errored out with exit status 1: command: /home/ubuntu/anaconda3/envs/ztf/bin/python /home/ubuntu/anaconda3/envs/ztf/lib/python3.6/site-packages/pip install --ignore-installed --no-user --prefix /.