Steps:

1. Enable Audience Targeting:

   • Go to Site Settings → Site Pages → click the gear icon → Library Settings → enable Audience Targeting under the "Audience targeting settings".

2. Target the Content:

   • In a web part like Highlighted Content, turn on Audience Targeting under the web part settings.

   • Assign one or more Microsoft 365 groups to a page or content item.

Please change your code to this:

@Query("select i from Brand i where i.name like %:name% and i.reviewStage = :stage")
Page<Brand> getLikeName(@Param("name") String name, @Param("stage") ReviewStage stage, Pageable pageable);

or this one:

@Query(value = "select * from brand i where i.name like CONCAT('%', ?1, '%') and i.review_stage = ?2", nativeQuery = true)
List<Brand> getLikeName(String name, String reviewStage, Pageable pageable);

You should use Wisper.cpp and not Python wisper
Bcoz if you are trying to work in Mobile devices Wisper Python wont work
We are currently implementing in my DD-DigitalDiary project

Last comment of Remy Lebeau for previous answer worked to me - "don't declare the ifstream as a global variable". I just made a few local variables like

ofstream flog;

Hint: Maintain a backup of your project either locally or as repository before making changes. Is simple then to revert or compare, Nb I use BeyondCompare for comparing code versions locally.

I also made a module called jcon

https://pypi.org/project/jcon/0.5.1/

thanks for the update, that makes more sense now. Just to check, when you say the vessel moves to the seized resource unit, do you mean it goes to the home location of the berth (like BDrop1, BDrop2, etc.) that's linked to the resource?

Also, does the Seize block automatically make the vessel wait if all the berths are full, or did you have to add anything extra for that? Just trying to set up something similar and want to make sure I’m on the right track.

give your code about when and where to run AppLinkData.fetchDeferredAppLinkData

I found the file where throws the error (No existing trace found).

• https://github.com/openai/openai-agents-js/blob/main/packages/agents-core/src/tracing/context.ts

After I installed `@openai/agents-core` package, for I want to try some logic, everything start working and the error gone away.

I don't know why but all I did is "I installed `@openai/agents-core` package".

The problem was solved. thank you.

Your styles aren't loading because Next.js thinks they’re not being used and skips them.

Just import your CSS file directly in your app, like in layout.tsx:

import 'component-library/dist/main.css'; 

I encountered the same issue on my DB,

Found the solution to trigger the pruning is doing the conversion using ::timestamp without time zone or with no conversion at all, at planning time postgres is not recognizing the index nor the pruning with different data type

Same issue was happening to me.

I got it fixed by updating macOS, updating Xcode to latest versions,

and re-running the following:

flutter clean 
rm -rf ios/Pods ios/Podfile.lock ios/.symlinks ios/Flutter/ephemeral 
cd ios && pod install --repo-update && cd ..

Failed to load System.Private.CoreLib.dll (error code 0x80070002) Path: C:\Program Files\dotnet\shared\Microsoft.NETCore.App\8.0.14\System.Private.CoreLib.dll Error message: Could not load file or assembly 'C:\Program Files\dotnet\shared\Microsoft.NETCore.App\8.0.14\System.Private.CoreLib.dll'. The system cannot find the file specified. (0x80070002) Failed to create CoreCLR, HRESULT: 0x80070002

Deleted LaunchScreen from the project (Remove reference) and add it back in solved it. Restarting device or reinstalling didn't work.

To do this for multiple files use the command:

for file in *.tif; do

tesseract "$file" "${file%.png}" --psm 6 lstm.train

done

This will generate all the lstm files

Use api method from postman

Endpoint take from api_server of your clearml config. For example

https://your_url:443/queues.update

And write body request

{

"name": "default",

"system_tags": ["default"]

}

Do not forget Authorize (create Bearer token, from method auth.login)

Literally wasted 12 hrs in others ai platform.just randomly googled it & here problem solved🫰.stackOverflow wont die

In the .env file try this way
MONGODB_URI = '<your URL'
....
It work for me

I am having the same problem when using stumpwm on arch-linux

I have eDP (laptop) and monitor (HDMI-A-0). I prefer having the monitor to left-of laptop, and things were working fine until yesterday when I connected a projector and somehow things got messed up and I'm unable to recover now. Now, when I login to stumpwm, the default mode is monitor to right-of laptop, and things work fine. But when I move the monitor to left-of laptop,

1. xrandr shows correct values set. even the background image takes the full screen

2. however, the windows only take up partial screen on monitor. and the windows on the laptop overflow onto the monitor

I've tried, arandr, but get the same results

You can check the mapbox option, have free license quota based on number of users or API usage, more info in their web:

https://www.mapbox.com/pricing

I think the "clunky, non-functional approach" is the best you currently have available. It's clear and to the point.

It depends.

Generally speaking, your learning rate is independent of batch size if you divide loss by batch size, otherwise you have to adjust lr when batch size changes.

You can extract each frame from the input video and identify white background pixels using color thresholding in the HSV color space. and then create an alpha channel is by marking the white regions as fully transparent and the rest as opaque. Save these processed frames are PNG images with RGBA channels. Finally, encode the sequence of transparent PNG frames into a .webm video using the VP9 codec with specific FFmpeg settings that preserve transparency, resulting in a video where the original white background has been removed and replaced with true transparency. You can refer to: https://colab.research.google.com/drive/1Jz39wRN4hiJbvGsrOTFYP4uymjhOlJgF?usp=sharing

• Port mapping ok.

• default root user use "%" with mysql and mariadb.

• Don't work : "localhost", "127.0.0.1", docker inspect IPAddress (ex. 172.19.0.3), network container aliases (OP "msqli"), etc, etc...

  So the solution was to use "host.docker.internal" !

Hi I made this because react-native-datetimepicker doesn't work. Please feel free to try it.

https://github.com/sugitata/react-native-year-month-picker-select

**                                     **           **     **  
                  **   **     //                         **   **     //           /**    /**  
 ******  ******  //** **       **  *****  ****** ****** //** **       **  *****  ****** ******
//**//* **////**  //***       /** **///**//**//*//**//*  //***       /** **///**///**/ ///**/ 
 /** / /**   /**   /**        /**/******* /** /  /** /    /**        /**/*******  /**    /**  
 /**   /**   /**   **       **/**/**////  /**    /**      **       **/**/**////   /**    /**  
/***   //******   **       //*** //******/***   /***     **       //*** //******  //**   //** 
///     //////   //         ///   ////// ///    ///     //         ///   //////    //     //

With mypy, you can put # type: ignore at the very top of the file (after shebang and encoding declarations, afaik) and it will not type-check the whole file. This is probably not advisable and this issue probably indicates you’re doing something wrong. From where are you importing AppConfig? Although it may just be that your linter or type-checker is misinterpreting literals as a separate type…

In macbook air M1 fixed by

sudo codesign --force --deep --sign - /Applications/Visual\ Studio\ Code.app

You probably want to set up an AWS IAM user that has permission to your AWS Cognito. Then your AWS IAM user can generate an access key to use in n8n.

https://docs.n8n.io/integrations/builtin/credentials/aws/#using-api-access-key

OAuth is more for authenticating multiple other users. For example, say you had many users and they each had different permissions. In this case, you might need OAuth so that the user HAS to sign in and you can only do what they have permissions to do.

def chapter_body(self, text):
    self.set_font("Arial", size=12)
    # Encode text to avoid Unicode errors
    text = text.encode('latin-1', 'replace').decode('latin-1')
    self.multi_cell(0, 10, text)

As per Maxim's answer, but the link died. Add this to the config file:

<system.net>
    <mailSettings>
        <smtp>
            <network
                clientDomain="mail.domain.com"
            />
        </smtp>
    </mailSettings>
</system.net>

Alternatively, use something like Lowcoder to build a UI that manages and triggers your workflow.

https://docs.lowcoder.cloud/lowcoder-documentation/workflows/n8n-integration

If all you want is start/stop/restart/status then:

pip install db4e-systemd

Documentation

Unfortunately,

: myexp ( u1 u2 -- u3 ) \ u3 = u1^u2   over swap 1 ?do over * loop nip ;

Hangs if the exponent is 0. The correct answer is 1.

: myexp ( u1 u2 -- u3 ) \ u3 = u1^u2   1 swap 0 ?do over * loop nip ;

Will fix that, but it hangs if the exponent is negative.

I think this Pull-Request can meet your need: https://github.com/elastic/elasticsearch/pull/130054

Just add this and try to login. On Github Desktop app in File > options> Advanced:

Something like this should work

%%{init: {'gantt': { 'leftPadding': 100 }}}%%

For anyone else suffering with this issue but still getting stuck on connection attempts even after installing the server under ~/.vscode-server/bin/{commit-hash}, there is a new setting in the remote-ssh extension for getting past the connection attempts with Microsoft's update server.

Near the bottom of the extension options there's a tick-box that says 'Use Exec Server' (Uses the new bootstrapping mode when connecting to a server. Can be toggled off in the event of connection issues."

Untick this setting(Use Exec Server) box and the connection can be made without trying to download and install all the packages first. This will require the grunt-work from the previous answers in installing the package manually but will also bypass the need to be connected to the internet and having it fail because the local/and remote server are both offline.

Good luck!

You can just apply a fixed random pattern of sign flips to the input data and then do a fast Walsh Hadamard transform.

Repeat that again for sparse input data.

https://sites.google.com/view/algorithmshortcuts/home

Code:

https://archive.org/details/software?tab=collection&query=mini+java+collection&and%5B%5D=year%3A%5B2025+TO+2025%5D

If you want to be minimalist:

print(*range(0,101,15))

First try restarting your IDE and see if something happens. If not then Delete your python installation using revo uninstaller to ensure deep clean. Reinstall python using the official installer from their website. Make sure to use the default installation and add it to PATH. After that create a .venv in your project directory. If you already have one delete it. And reinstall all the packages in the venv. I have faced issues with pyside6 similar to yours where python can't find a package even though it's clearly installed. The only fix I have found is to delete the .venv folder and reinstall everything. It's apparently a common-ish thing with pyside6.

It's common sense to not use any SDK installed from the ms store. MS store is good for regular people using regular apps not Devs using dev tools. Btw, use pycharm for easier python coding. Pycharm CE is free.

try which conda activate
you could get a dir like this /usr/local/Caskroom/miniconda/base

/{base} is the conda venv name,
change this with targeted env
then you could activate the env from it at /bin/activate or with activate script dir

Did you manage to solve it by any chance?

That method failed for me because the "copy" had hidden info regarding begin and end of cell. As a result, I could not properly parse the data as a text string with End of Line commands. I am looking into the HTMLFile object solution.

The relevant section of code should be modified as follows:

<MudDataGrid Class="table" Items="context.Vendors"  >
    <Columns>
        <MudBlazor.PropertyColumn Property="@(vendor => vendor.Name)" />
        <MudBlazor.PropertyColumn Property="@(vendor => vendor.Description)" />
        <MudBlazor.TemplateColumn Title="Action" Context="vendor">
            <CellTemplate>
                <MudLink Href="@($"vendors/edit?id={vendor.Item.Id}")">Edit</MudLink>
            </CellTemplate>
        </MudBlazor.TemplateColumn>
    </Columns>
</MudDataGrid>

My chief errors were not using CellTemplate, and not understanding .Item.*

In Strapi 5, the process is the same as Martin Larizzate described, but the property names are different:

files: {
  filepath: filePath, 
  originalFilename: fileName,
  mimetype: mime.getType(filePath),
  size: stats.size,
}

^Run this command in the terminal this will fix the issue
flutter create --platforms=ios .

If you want to be minimalist, this gets the job done:

i=0
n=0
while n!=100:
  i=i+15
  n+=1
  print(i)

nvm, I'm just dumb and didn't notice the terminal is interactive... I could basically select a running target and ping logs from it in the window

Regex can do the trick. Note that I've used the "s" flag, so a "." includes new lines - normally the first capture group would start at the beginning of the current line, rather than the beginning of all text:

echo preg_replace('/.*(?<=<body>)(.*)(?=<\/body>).*/s', '$1' ,$data); // get body text only

I realised I could add a CHR(34) instead of double quote and it worked.

There is a direct way of adding sources, resources, headers from folders nested in a Visual Studio projects directory.

In the toolbar of Solution Explorer, click "Show All Files".

Next right click the folder you need and select "Include in Project".

Now Visual Studio will not automatically scan this folder again and again. So we have two kinds of options.

1. If you are moving sources and headers from somewhere to these folders, just include the folders once again and Visual Studio will add them. This will however include all files in the folder that you manually removed from the project yourself.

2. If you want to create a new file in that folder from Solution Explorer, either

   1. In the "Show All Files" view, add by right clicking on the folder and selecting Add option.

   2. In the Default view, include the path along with the filename when select Add option. So instead of entering MyFile.cpp, it goes src/MyFile.cpp. Visual Studio will create any required folders in there is not.

For anyone having same question as I: "Obi softbody" in the unity asstestore can do the trick for both physics and visual.

But if you want something more visual and not physically so much effective "magica cloth" is the tool; it is also on the unity asstestore

There are some other tools as soft mesh deformer...

Can you provide context such as what operating system you're using, and what application you are using to write/execute code? My best guess is that the file browser opening in one case is the code editor's file explorer built into the app.

HTML is markup not a programming language . Use an actual programming language like PHP, JavaScript etc. There are many free resources for learning any of these languages and also for how to use it in raspberry pi. I would recommend JavaScript with NodeJS.

Found the solution by changing the format of table. Thank you, community!

Adding additional policies to the IAM role still did not work for me. This video got me the basics https://www.bing.com/videos/riverview/relatedvideo?&q=step+by+step+to+configure+RDS+SQL+restore+from+S3&&mid=C387ABE5AC396000E939C387ABE5AC396000E939&&mcid=507E1C5F5F944F0BB9504BE3A01DA122&FORM=VRDGAR

But it was still failing until I reset the master rds password!

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

Dinh Phuong Linh la culi cua a ma sao cam r kia 😂🍼🍼

for (int i = 6; i>0; i--){
        cout<<space;
        for (int j = i-1; j >=0; j--){
            cout << star << " ";   
        }
        space += " ";
        cout << endl;
    }
this works if you're trying to create a triangle that tapers downwards

I suggest changing the name to The type or namespace name could not be found

From your comment, I see that this is because the assembly did not load correctly.

This is an answer for anyone else having this problem - restarting Unity and your code editor should help, if not, check if you see miscellaneous files next to your file name (where the assembly name should be) - if it does - there are many ways to solve the miscellaneous files problem online

Three things are involved here. The tmh files are created by the WPP process (part of Tracing), and it appears WPP won't run if the C build doesn't complete (compile and/or link errors). Lastly, you will need include statements like #include "SourceFile.tmh" in your code. If a .C file includes both an A.h and A.tmh, then A.h should be first. If the C code finally builds, then WPP should run and create all the .tmh files. The next build should be clean, with no complains on missing .tmh files.

The WPP settings are located inside the project's properties dialog. A Visual Studio driver project will have a "Wpp Tracing" entry on the left. Go to that page, and make sure "Run Wpp Tracing" is turned on ("Yes").

if you check obj type of the four digit number that is input e.g. print(type(number)) you will find it is a string. Which should help you find the issues with your code.

If someone is still facing this, add to the esbuild config:

  exclude:
      - 'jsdom'
      - 'canvas'
  external:
    - 'jsdom'
    - 'canvas'

It needs to be both to work! Workaround discussed here:

https://github.com/evanw/esbuild/issues/1311

Have you tried defining the table as it is and then using :last-child?

table.rti.app5 tbody td:last-child {
  text-align: center !important;
}

Do you have any more information you can share as to what you have tried, what more of your CSS looks like and where / how it is used

You can find the pip config location by running pip config list -v

browse for

C:\ProgramData\pip\pip.ini

C:\Users\<user>\pip\pip.ini

remove nvidia url and set below config

no-cache-dir = true
index-url = https://pypi.org/simple

trusted-host =
               pypi.org/simple
** post which restart the terminal

extension Data {
    func base64URLEncode() -> String {
        let base64 = self.base64EncodedString()
        let base64URL = base64
            .replacingOccurrences(of: "+", with: "-")
            .replacingOccurrences(of: "/", with: "_")
            .replacingOccurrences(of: "=", with: "")
        return base64URL
    }
}

After adding lines that were described here I still received an error.

I dig somewhere that in fact you need to delete some of them.

For me it was failing when I had a config:

  "compilerOptions": {
    ....
    "module": "es2022",
    "lib": [
      "esnext",
      "esnext.asynciterable"
    ],
    "allowSyntheticDefaultImports": true
  },

I need to remove lines:

"esnext.asynciterable"
"allowSyntheticDefaultImports": true

Then command added back the missing items and succed.

The reason you see that is because you're running it in an interactive shell.

Try this

npx -y pnpm install --silent

This pre-confirms the question.

options.addArguments("--force-device-scale-factor=0.65");  // Apply 65% zoom in headless
           

I'm still facing this issue and haven't found a working solution yet.

Every time I run my Flutter app on iOS Simulator, I get the following error:

Error initializing DevFS:
DevFSException(Service disconnected, _createDevFS: (-32000) Service connection disposed, null)

Steps I've tried so far:

• Ran flutter clean, flutter pub get

• Reinstalled pods (deleted Pods, Podfile.lock, .symlinks, etc.)

• Rebuilt iOS project with pod install

• Erased and restarted all simulators

• Simplified my main.dart to just show a Text widget

• Made sure all xcconfig files include CocoaPods settings

• Tried different iOS simulator devices

• Updated Flutter to latest stable (Flutter 3.22.x at time of writing)

• Verified flutter doctor has no errors

• Also tried flutter run --release

Still getting the same error.

Full crash report shows a segmentation fault (SIGSEGV) around UIImageView/intrinsicContentSize which may suggest a corrupted image or native issue.

If anyone has encountered this or found a workaround, I’d really appreciate any tips or directions

As far as I understand and use in projects,

- Service layer deals with business logic and might returns business entities
- Controller layer might deal with what it gets and outputs to the outer world (DTOs)

Like imagine a scenario where:
- A service might need to talk to another service, why they should deal with DTOs?
- You may have different api versions and controllers that would (may do some extra work) but also return different response shape based on new or different DTOs, so would it be wise to refactor (or even touch) service layer in that case to adapt to this change?

YOU CAN APPROVE YOUR OWN PRs. here is how:

go to the go into the repo -> setting -> rules -> rulesets -> create a new ruleset, and set Bypass list, and then add your roles: Repository admin, or what role you want.

-if you have not set yourself as the admin, then at the repo ROOT -> Organization Roles -> Role Assignments, then assign yourself admin, or whatever else

As of iOS 17, the current way to set the badge number is by calling:

import UserNotifications
let newBadgeCount = 5
UNUserNotificationCenter.current().setBadgeCount(newBadgeCount)

On one of my tables, setting columnMenu='legacy' stopped it from hiding on hover but not sure why. version 32.3.7

Looking at your code, I can see the issue. The problem is in your backend uploadImage function. You're checking for request.body.file but when using multer with file uploads, the file information is stored in request.file (singular), not request.body.file.
Your backend controller is looking in the wrong place for the file info.

This was a Firefox bug that has now been fixed as of Firefox 140. The code in the original question now works as expected across browsers.

Certain vscode extensions do stuff like that. Usually uninstalling and reinstalling the extension works. If that doesn't work disable every other extension. If you can't disable an extension uninstall it. Then check if your extension that was causing the issue is working or not. If that doesn't work then your vscode installation is either broken or your Windows OS has some issues. Stuff like this has happened to me before twice, though not with a Debugger.

Maybe my answer doesn’t cover all your questions, but here’s what I’ve found.

The go list command shows which packages will be built( build list). When you run go list -m all, it shows the modules really loaded during the execution —check the official reference for details on how it works.

Finally, go download fetches every dependency declared in your go.mod and it applies to all dependencies of the main module; see the Go documentation for more information.

Not sure what was going on but I would do File > Open Folder and the contents of the folder would not load in solution explorer. I've gone through all the steps in this thread and only this one worked but it wasn't very clear.

Angad Kumar put me on the right path, but wasn't 100% based since the instructions were not perfect.

I deleted everything found in %appdata%\Microsoft\VisualStudio and that resolved the issue.

The same code is treid to run with the help of gemini and am able to compare the code with out any differences between manual run and Gemini run and with exact replica can run the code on Gemini without issues.

@Retryable annotation has a "recover" parameter that can be used to define the exact recovery method regardless of the exception class or method signature.
example:

@Retryable(retryFor = {HttpServerErrorException.class, ResourceAccessException.class}, recover="myRecover")

As I mentioned in the comments, Apple took exactly 30m to finish the process. Most probably they added some cool off time.

Looking to refresh your space without spending a fortune? Tips DrHomey shares smart, budget-friendly ideas to decorate your home with style and personality. From repurposing old items and adding indoor plants to using lighting, wall art, and DIY crafts, these decorating tips help create a cozy, functional atmosphere. Whether it’s your living room, bedroom, or garden area, DrHomey’s advice focuses on simple upgrades that make a big impact. Perfect for renters, homeowners, and anyone wanting a home that feels uniquely theirs.

This feature depends on the Analyze model-driven apps and Microsoft Dataverse telemetry with Application Insights feature.

The instructions describe how to enable exporting data to Application Insights. See Export data to Application Insights

Step 6 says:

Select the Azure subscription, resource group, and Application Insights environment that you're exporting data to. Select Next.

You must have contributor, writer, or admin rights to the Application Insights environment. Typically, one production environment or tenant maps to one Application Insights environment.

You should verify this is set up correctly and that you have the appropriate rights to the Application Insights environment.

You should not have to set the organization.telemetryinstrumentationkey column. That column is valid for update but not valid for read. It should never return the key value. This is by design. But it makes it hard to verify whether the key value is stored.

const transporter = nodemailer.createTransport({
  host: "smtp.gmail.com",
  port: 587, // 🔁 Use 587 instead of 465
  secure: false, // 🔁 false for STARTTLS
  auth: {
    user: process.env.EMAIL_USER,
    pass: process.env.EMAIL_PASS,
  },
});

I have use this ,and solved my problem

6767557676467557756657656$%%%-$-&8:'*&"

ssh is a text only solution (a terminal emulator with encryption). Plain terminal sessions can't do GUI...

Just use Pylytic, an alternative to built-in eval() for evaluating complex mathematical expressions

i found an article that explains the processes and lists a tool called MacSymbolicator that makes this way easier: https://www.avanderlee.com/xcode/symbolicate-crash-logs-reports/

The problem was with my validating webhook where for Delete call I was taking req.Object.Raw instead of req.OldObject.Raw.

The Goodwill Outlet, often referred to as "The Bins," is a chaotic, exhilarating, and sometimes overwhelming treasure hunt where merchandise is sold by the pound rather than at fixed prices. Unlike

I created a rule set that did a url redirect which seems to work:

IF Condition Host name Equal api.mydomain.de

THEN URL redirect Permanent redirect (308) match request apim-name.azure-api.net

were you able to fix this issue? I am having the same problem.

What happened in my scenario was that we have a connection with no limit that is driven off a variable. The variable was pointing to a server that was removed and no longer on our network. Once we fixed the variable the problem was resolved.

Really a silly mistake

solved here: https://github.com/tauri-apps/plugins-workspace/issues/2798

shift + enter renders the markdown cell in vs code. double-clicking it reverts back.

What happen if i used the version "apexcharts": "^3.50.0" and "@angular/core": "^17.3.11", ??

there will be some error ?

Here's a tutorial using Slack to create a custom function using type script. This covers all the steps to create the app, get the PAN, and deploy the code. Specifically this is with respect to Github API as that's what I'm implementing now https://tools.slack.dev/deno-slack-sdk/tutorials/github-issues-app/

NVIDIA GPU Installation steps and guide (Windows 11):

----------------------------------------------------

Step 1: Download and Install ==> "cuda_11.8.0_522.06_windows"

Make Sure Installed version correct or not using this cmd in powershell =\\\> "nvcc --version"  

**Make Sure adding the Environmental Variable Path:**  

    **User Variable:**  

        CUDA_PATH =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8"  

        path =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\bin"  

                "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\libnvvp"  

    **System Variable:**  

        CUDA_PATH =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8"  

        CUDA_PATH_V11_8 =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8"  

        path =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\bin"  

                "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\libnvvp"

Step 2: Download and Extract ==> "cudnn-windows-x86_64-8.6.0.163_cuda11-archive"

Copy ==\\\> (bin, include, lib) files paste it in below paths  

for bin =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\bin"  

for include =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\include"  

for lib =\\\> "C:\\\\Program Files\\\\NVIDIA GPU Computing Toolkit\\\\CUDA\\\\v11.8\\\\lib\\\\x64"

Step 3: Download and install ==> "Python Version: 3.10.0"

Make Sure adding the Environmental Variable Path:  

    **User Variable:**  

        path =\\\> "C:\\\\Users\\\\{System Name}\\\\AppData\\\\Local\\\\Programs\\\\Python\\\\Python310\\\\Scripts\\\\"  

                "C:\\\\Users\\\\Balaji P\\\\AppData\\\\Local\\\\Programs\\\\Python\\\\Python310\\\\"  

      

    **System Variable:**  

        (not need)

Step 4: install using powershell ==> "pip install tensorflow==2.13.0"

Make Sure Installed version correct or not using this cmd in powershell =\\\> "pip show tensorflow"

Step 5: install using powershell ==> "pip install tensorflow_gpu==2.10.0"

Step 6: Make activate script using powershell (Enter First)==> cd "C:\{Your Project}"

                                    (Once Done above Enter Second) ==\\\> ".\\\\.venv\\\\Scripts\\\\activate"

Final Step: Run Below code check GPU showing or not

    import tensorflow as tf

    import sys



    print(f"Python Version: {sys.version}")

    print(f"TensorFlow Version: {tf.\__version_\_}")

    print("-" \* 30)



    \# This is the most important part

    gpu_devices = tf.config.list_physical_devices('GPU')



    print(f"Num GPUs Available: {len(gpu_devices)}")



    if gpu_devices:

        print("Found GPU(s):")

        for gpu in gpu_devices:

            print(f"  - {gpu}")

        print("\\nTensorFlow is built with CUDA:", tf.test.is_built_with_cuda())

        print("TensorFlow is built with GPU support:", tf.test.is_built_with_gpu_support())

    else:

        print("No GPU was detected by TensorFlow.")

        print("Please check the following:")

        print("1. Is the NVIDIA driver installed and working? (Run \`nvidia-smi\` in your terminal)")

        print("2. Is the correct version of CUDA Toolkit installed?")

        print("3. Is the correct version of cuDNN installed and its files copied to the CUDA Toolkit directory?")

        print("4. Do the TensorFlow, CUDA, and cuDNN versions match the official build configurations?")

        print("   (Check https://www.tensorflow.org/install/source#gpu)")

        

        

    #Success Output below:

    Python Version: 3.10.0 (tags/v3.10.0:b494f59, Oct  4 2021, 19:00:18) \[MSC v.1929 64 bit (AMD64)\]

    TensorFlow Version: 2.10.0

    ------------------------------

    Num GPUs Available: 1

    Found GPU(s):

      - PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')



    TensorFlow is built with CUDA: True

    TensorFlow is built with GPU support: True

Make Sure check with versions using run below cmd in powershell:

"nvcc --version"

"pip show tensorflow"