2 days ago I also created a chrome extension, which can apply various dark mode colors and text color to any website. I created that to help me in choosing dark mode colors for my websites. With the help of chat gpt it took me around 30 minutes.but I didn't publish it as it requires money to publish the extension. So I can only use it myself 😅
It seems like the implementation might not be optimal (AI part). Providing the entire context, such as your very Big prompt and DOM information, to GPT every time could be costly in terms of input tokens.
@@codecurrents can we do encoding to decrease token size??? is it any way possible or like other way: to remove unwanted or repetitive characters like 'is', 'a', 'the', kinda words... or using text analyser to remove them using python module or normal filter condition; one other possible way like to reduce prompt input characters such that it doesn't loose semantic of it (as token size depends on characters written, that's why got these ideas randomly, hit me up if you like so)
It would be more interesting if we could find a way to use open-source models, as many people, including me, don't have an OpenAI API key. I usually work with open-source models, so I'll try to explore this myself. Others can also work on this-it would be a really interesting problem to solve. Piyush bhai, you’ve always been my go-to person for learning something new. Much love to you!
You can try free APIs like: 1) Groq - Similar format to OpenAI's API, making it a drag-and-drop replacement. 2) Gemini - Has a different structure, requiring some additional code. Alternatively, you can use Ollama to run open-source models locally and access its API. However, I’m not sure if Ollama’s API can be accessed directly from the browser. All the best
Hi Piyush, I am trying to build chrome extension with same approach where a user ask a query from text box but I am getting error as 429- openAI rate limit Do I need to have a chatgpt subscription ??
Sir, I have a question. For each user request, we are sending a lengthy prompt, which increases token usage and gradually raises costs. Can we optimize this by providing only the necessary context for the GPT model, perhaps by using a Retrieval-Augmented Generation (RAG) model or a similar approach to save costs effectively?
Most chatbots store full chat history. RAG is different - it pulls in outside info for better answers. It works well for research, summaries, and specialized tasks. Trying RAG in regular chats could be an interesting experiment.
i thought you made this extension with the new AI features which are getting available in the new Web API i would really appreciate if you make a video on that
Sorry, but not able to run project locally, getting a blank screen when i run npm run dev. I Installed the extensions. It's good to include to steps to run project locally or make a small video how to run locally
can be learnt and it's like saying, "a website can't be made in rust ?" well it can be, but who F has that much of energy and will power to do, and fine-tuning a neural network is already hard enough in python + it takes lots of money and resources, nobody is ready to do it, its more like a scale ⚖️, one side is time in development and the other side is time consumed in generation, to increase one side, other side have to go down, if want fast Ai we can make it in rust or javascript, but then the time taken to develop that Ai will too much, and in this money making industry investors wants fst return.
You are my inspiration. How can I be a very good software engineer just like you. My only goal is to be best Software Engineer. Can you provide some tips???
Your all video are helpful But this is insane if can do obviously you can Make a video on aws livefaceness detection without using amplify plz do ( anti spoofing )
Impressive work, bhai! There is a better way to find the selected language and current code in the leetcode problem page. We can access monaco editor instance. With that instance, we can get those. const editor = window.monaco.editor.getModels()[0] const language = editor.getLanguageId() const code = editor.getValue() [I tried window.monaco.editor.getModel() but it's not working, I am not sure why. So I had to use first element of the array of window.monaco.editor.getModels() 😥]
Maza agaya 🫶
are sir app yahaa pee
Angular try Karo paaji aur bhi maza ayega 💀
Congratulations, Piyush, for building it. And, thanks for Inspiration.
Very Lucky to have a Gem like Piyush Garg for beginners like me.
14:17 you can use XPATH to select that button and extract its string which has selected programming language mentioned in it
Amazing stuff
Loved it ❤❤
Thank U Bhaiya I am following you the journey of coding 😊
I am working on the exact same idea. Happy to see you got the same idea too.
I already have a leetcode video solutions extension with 30K+ users.
from where you get the video of solution ? did you make the video yourself? or get from somewhere else
Name of the extension please say
@@Rajeshraju-n8u just google "leetcode video solutions".
Dsa folks: I will solve leetcode hard problems today.
Le developers: Let me help you with that😂
If this extension requires a server, hosting costs make it unsustainable without charging users. Great project though-looking forward to its progress!
You are greate man. love you buddy from the depth of my heART
First time commenting on your video, lot's of respect to you by the way nice video/information ❤
That is amazing project just in 45 minutes.
Bro your content 🔥🔥 🚀🚀
Awesome bhaiya 🔥🔥🔥🔥
JavaScript is beautiful ❤️❤️❤️❤️
I will try it out and try to contribute too 🔥🔥💯
2 days ago I also created a chrome extension, which can apply various dark mode colors and text color to any website. I created that to help me in choosing dark mode colors for my websites. With the help of chat gpt it took me around 30 minutes.but I didn't publish it as it requires money to publish the extension. So I can only use it myself 😅
Nice bro, can i get this project repo please
Can I get the repo link if possible bro ?
Hey can you drop me the repo for it or the dist folder?
Excited
this is what I need right now.
great video and nice work
Wonderfull creation!
It seems like the implementation might not be optimal (AI part). Providing the entire context, such as your very Big prompt and DOM information, to GPT every time could be costly in terms of input tokens.
Exactly, can you help me to make it better? You can pitch some good ideas as well.
@@codecurrents can we do encoding to decrease token size??? is it any way possible or like other way: to remove unwanted or repetitive characters like 'is', 'a', 'the', kinda words... or using text analyser to remove them using python module or normal filter condition; one other possible way like to reduce prompt input characters such that it doesn't loose semantic of it (as token size depends on characters written, that's why got these ideas randomly, hit me up if you like so)
great sir ,thanks❤❤❤❤❤❤
Whenever I work on an idea, it goes well for some time.
But then, it eventually gets stuck in errors.
I also created a chrome extension Detoxify which filter you tube search videos according to views, comments and likes
please share me the github link
Can you share GitHub link for that project..
used web scraper for retreiving info??
It would be more interesting if we could find a way to use open-source models, as many people, including me, don't have an OpenAI API key. I usually work with open-source models, so I'll try to explore this myself. Others can also work on this-it would be a really interesting problem to solve.
Piyush bhai, you’ve always been my go-to person for learning something new. Much love to you!
You can try free APIs like:
1) Groq - Similar format to OpenAI's API, making it a drag-and-drop replacement.
2) Gemini - Has a different structure, requiring some additional code.
Alternatively, you can use Ollama to run open-source models locally and access its API. However, I’m not sure if Ollama’s API can be accessed directly from the browser.
All the best
I am ready for it❤❤❤❤❤❤
Hi Piyush,
I am trying to build chrome extension with same approach where a user ask a query from text box but I am getting error as 429- openAI rate limit
Do I need to have a chatgpt subscription ??
yup, use gemini instead of openAI
I will try to contribute in this extension
Sir, I have a question. For each user request, we are sending a lengthy prompt, which increases token usage and gradually raises costs. Can we optimize this by providing only the necessary context for the GPT model, perhaps by using a Retrieval-Augmented Generation (RAG) model or a similar approach to save costs effectively?
Most chatbots store full chat history. RAG is different - it pulls in outside info for better answers. It works well for research, summaries, and specialized tasks.
Trying RAG in regular chats could be an interesting experiment.
@@sayantanbiswas9702 yes I think so .
Is the process same for different bundlers or some minor changes?
16:48 .. thanks for the key 👍
How to get the key
Awesome
Hi piyush, hope you are still active in this project as no commits are done from the last 2 weeks.
i thought you made this extension with the new AI features which are getting available in the new Web API i would really appreciate if you make a video on that
Next video please on School Management System using ReactJS
bhai aapne abhi tk monetize kyo ni kiya apna channel
Piyush please guide in proper way like what to we have to install and other things.
Hey piyush can you teach us Goland?
can we cheat in live contest using this?
Yes you can. But be ready to lose the account when someone reports plag 😊
can we do something like download ollama in our system and then use it using extension locally and genrate code, like use ollava model or something
Ja rha hu readme file se space hata ke kontributor ban jaunga 🤡
😂
Shall we use Gemini ai api ?? For free usage
Sorry, but not able to run project locally, getting a blank screen when i run npm run dev. I Installed the extensions. It's good to include to steps to run project locally or make a small video how to run locally
Is there any step by process how to get OpenAI key
openai api is paid or free ?
can we get the video on ,How to contribute in gsoc project for begginer
extension name plz
❤❤❤
Bro is killing us with this banger content❤😭😭
Wow i didn't know building Chrome extension is this simple. I mean if you already have knowledge of web dev
Ai cant be learned in JavaScript language?
can be learnt and it's like saying, "a website can't be made in rust ?" well it can be, but who F has that much of energy and will power to do, and fine-tuning a neural network is already hard enough in python + it takes lots of money and resources, nobody is ready to do it, its more like a scale ⚖️, one side is time in development and the other side is time consumed in generation, to increase one side, other side have to go down, if want fast Ai we can make it in rust or javascript, but then the time taken to develop that Ai will too much, and in this money making industry investors wants fst return.
Har din ise me Lage rahate ho ya gindigi bhi gite ho
Hi, Can u make a tutorial on how to make Safari Extension without Ai.
Outstanding yaar ap is ma Gemini api key function bi add kara
There’s no way I can make it in 40 mins 😂
Next Video : I built My Own Girlfriend 👧
I actually need this project fr
Already made by developer Rahul... But that's not open-source...
I need this too
Ban gya cool?
Much needed
You are my inspiration. How can I be a very good software engineer just like you. My only goal is to be best Software Engineer. Can you provide some tips???
hello sir give me some chrome extension idea
Bro Just want to have a conversation with you can you do that
me bhi bna rha hun sir 🫡
But video to 18min ki hi hai
If you speaking another language please put in title
First ❤
that's great awesome if you need my help i will be touch for my better understanding and contributing
We shouldn't promote the use of AI on leetcode...
First one
I uploaded the same video 😂😂 on my Channel lol Piyush bhai 🥳
Your all video are helpful
But this is insane if can do obviously you can
Make a video on aws livefaceness detection without using amplify plz do
( anti spoofing )
Impressive work, bhai!
There is a better way to find the selected language and current code in the leetcode problem page. We can access monaco editor instance. With that instance, we can get those.
const editor = window.monaco.editor.getModels()[0]
const language = editor.getLanguageId()
const code = editor.getValue()
[I tried window.monaco.editor.getModel() but it's not working, I am not sure why. So I had to use first element of the array of window.monaco.editor.getModels() 😥]
Own GF when 🤣🤣🤣🤣
@piyushGarg i put my apikey but its not workin g