2 ай бұрын
Пікірлер
@ee-lonlim8240 Күн бұрын
hi Jain, tks for this. can i seek your help on steps to deploy to front-end / web app - say via Azure web apps? tk u.
@eternal_dread666 2 күн бұрын
Can u do this with local LLM
@generativegeek 2 күн бұрын
it can be done. i will make a video on it in a few days. i am slightly tied up with lot of work these days.
@eternal_dread666 2 күн бұрын
@@generativegeek alright
@eternal_dread666 3 күн бұрын
I can’t seem to save the api key Says: API key save error
@generativegeek 3 күн бұрын
do you have write permission in the folder you are storing your .env file?
@generativegeek 3 күн бұрын
also please share link to a screenshot or your github repo. else its guesswork at my end ;)
@eternal_dread666 3 күн бұрын
@@generativegeek you mean I have to store the api key there?
@generativegeek 3 күн бұрын
in case of langflow you can create a variable and save your key there, like i did and use that.
@generativegeek 3 күн бұрын
check video from 14:12 onwards for steps.
@eternal_dread666 4 күн бұрын
so when i run the code it doesnt open the chatbot app whats wrong?
@generativegeek 3 күн бұрын
it will be great if you could share a small loom.com video or some screenshots. Its hard to debug without knowing more context. Please can you share your github repo with the code + some screenshots etc.
@eternal_dread666 3 күн бұрын
@@generativegeek alright do u have discord?
@eternal_dread666 3 күн бұрын
@@generativegeek do u have discord? I can add u and send u the errors
@eternal_dread666 4 күн бұрын
Can I do this with gemeni Ai? I think u gotta pay for OpenAI api
@generativegeek 4 күн бұрын
yes you can do this with Gemini as well. You can use the gemini embedding function. Watch one of the gemini video from the playlist to see how to use the gemini embeddings and store them in a datastore etc.
@eternal_dread666 4 күн бұрын
@@generativegeek THANKS do you know why i get this error when setting up virtual enviroment? ModuleNotFoundError: No module named 'pip._internal'
@darkmatter9583 6 күн бұрын
please my laptop is 300usd(pre covid19) ubuntu 4gb of ram i could buy a ram of 8gb and change it, even if is aws, or cloud is great for me can you give me advice please
@generativegeek 5 күн бұрын
you want to run these models on AWS? Sorry am not clear on the ask.
@darkmatter9583 5 күн бұрын
@@generativegeek i want to run it, doesnt matter if aws,local, google colab but my limitations are these one
@generativegeek 5 күн бұрын
@@darkmatter9583 got it. In that case i will recommend you can set up ollama on an aws or digital ocean linux instance. you will need to follow ollama linux instructions to install the service and model. Rest of the code will remain same. check ollama.com/download/linux
@eternal_dread666 8 күн бұрын
Why does my notebook code gives “false” instead of true?
@generativegeek 7 күн бұрын
it seems like your .env file is not in the right place and does it have the values as shown?
@eternal_dread666 7 күн бұрын
@@generativegeek how can I make in right place? I spent hours but didn’t find solutions
@generativegeek 7 күн бұрын
@@eternal_dread666 please send email with your steps to [email protected]
@eternal_dread666 7 күн бұрын
@@generativegeek actually fixed it I did 2 things Saved the .env file And installed dotenv in vsCode I didn’t u had to do those Actually Local LLM helped me with that
@eternal_dread666 6 күн бұрын
@@generativegeek i have another error when running second executable: NameError: name 'HuggingFaceHub' is not defined"
@gurerguncan 8 күн бұрын
Basically, what you did is, get a long video, divide it into short potentially viral videos. Am I right?
@generativegeek 7 күн бұрын
yes. thats exactly what was done.
@UshaJain-z8n 9 күн бұрын
Thank you for the video, I am facing an issue with Vector Search. I don't have an option on langflow for "Vector Search". Please help me with the steps.
@generativegeek 9 күн бұрын
which version of langflow are you on?
@vaibhav81 12 күн бұрын
0:00 - Introduction 0:08 - Nutrition Monitor Demo 0:25 - How it Works: Google Gemini Pro 0:50 - Project Setup: Streamlit 1:06 - Creating a Virtual Environment 1:36 - Installing Streamlit 1:56 - Environment Variables and Google AI 2:14 - Creating a Python Environment 2:47 - Verifying Installation 3:39 - Coding app.py 3:47 - Importing Libraries 4:37 - Creating .env File with API Key 5:19 - Main Coding File: app.py 5:31 - Page Configuration 6:59 - Loading Environment Variables 7:12 - Configuring Google Gemini AI 7:48 - Method: Get Gemini Response 9:49 - Handling Image Upload 11:25 - Returning Image Data 12:27 - Building the Sidebar 13:57 - Building the Header 14:33 - Displaying Uploaded Image 15:48 - Implementing Submit Button Logic 17:30 - Prompt Engineering: Input Prompt 19:19 - Input Button Code 20:51 - Code Completion and Execution 22:01 - Testing with Images 24:33 - Prompt Optimization: Table Output 24:55 - Conclusion and Outro
@Jaishreerammodijiom 15 күн бұрын
Bro can u share thr documentation
@generativegeek 13 күн бұрын
code is available at github.com/technofriends/nutrition-google-gemini-demo
@Jaishreerammodijiom 18 күн бұрын
Can u please give me the code i liked and subscribed you bro
@generativegeek 17 күн бұрын
thank you. code link is part of description.
@generativegeek 13 күн бұрын
code is here github.com/technofriends/nutrition-google-gemini-demo
@brianrowe1152 20 күн бұрын
Just asking. Why would I do this? When I can just prompt something like Meastro to just write the python code? The LLM is the interface now. I tried this and its neat, but at the end of the day, its faster to just ask Maestro to write the python. Not trying to start a riot, I just don't see why I would use this.
@generativegeek 7 күн бұрын
this is just a demonstration of what all is possible. the way to think about this is that if you know of the capabilities, you can orchestrate these type of pieces and build a solution.
@ShortsStudioOne 20 күн бұрын
Good tutorial, thanks if you have integrated the translation of subs in to desired language, this video would have been even better.
@generativegeek 20 күн бұрын
Thank you. You mean once we have the subs, add them to the video?
@generativegeek 20 күн бұрын
Please see this - kzfaq.info/get/bejne/fdSjltuavZ2vl3k.html 1. Use KZfaqTranscript API you can get the transcript. 2. Translate it into the language of choice 3. Use FFMPEG to build translated subs into the video.
@ShortsStudioOne 19 күн бұрын
@@generativegeek That is already covered in this video. I meant, after doing transcribe, translate the SRT to the language of our choice and add those translated subs to the video. Any ways, you got it covered in another video, I will combine both program in to one. I am using Tkinter to make a GUI for the same.
@asawarisatpute3634 20 күн бұрын
Thanks for this video, this is amazing video.I just have to know how to deploy it. I have locally installed langflow in my laptop and create a project which is successfully run. Now I have to make it everyone usable, can you please guide me or suggest me? Thanks in advance.
@explorewithrudransh8342 21 күн бұрын
Thank you. Loved the simplicity of this app. Learnt a lot.
@generativegeek 21 күн бұрын
Glad you liked it. god bless you.
@vinayakassajjanshetty7231 22 күн бұрын
really learned a lot and I think I have become the fan of you. Watching the videos are like tech movies. Moreover, I just want to say "Thankyou for efforts making me understand the core concepts".
@generativegeek 22 күн бұрын
thanks a lot for your appreciation. Your words mean a lot to me. Do share any concepts on which you want me to make upcoming videos.
@ayushmishra5861 22 күн бұрын
could you please use open source models instead?
@generativegeek 22 күн бұрын
sure. however google generative AI usage is also free. In the next video will use open source model from Hugging face.
@priyankasrivastava8356 24 күн бұрын
hello please can you add that extra bit of codes and update
@generativegeek 24 күн бұрын
yes the notebook is now ready. thanks for bringing it to my attention. Please share your email address for a Rs. 500 amazon voucher for being a helpful viewer of the channel. Notebook is at : colab.research.google.com/drive/1YP_zjPwudPydCii36K5_-fvtrV6L5nYF please share your email address at [email protected]
@mayankdwivedi4545 28 күн бұрын
Vector database Embedding
@generativegeek 23 күн бұрын
You got it right. Please share your email address for Rs. 200 voucher. Please send your email at [email protected]
@santiagomolinarestrepo6329 28 күн бұрын
Thank you so much for this. It helped me a lot.
@generativegeek 23 күн бұрын
i am glad you found it useful. thank you.
@sonaliagarwal8961 Ай бұрын
1. Vector Database 2. Embedding
@generativegeek Ай бұрын
You got it right. Please share your email address for Rs. 200 voucher. Please send your email at [email protected]
@vaibhavLehra Ай бұрын
1 vector database 2 embedding
@generativegeek Ай бұрын
You got it right. Please share your email address for Rs. 200 voucher. Please send your email at [email protected]
@jyotidwivedi8014 Ай бұрын
1. Vector database 2. Embedding
@generativegeek Ай бұрын
You got it right. Please share your email address for Rs. 200 voucher. Please send your email at [email protected]
@TryExcept-g5g Ай бұрын
1. Vector database 2. Embedding
@generativegeek Ай бұрын
you are correct. please share your email address by sending an email to [email protected].
@TechBaba-ch3pz Ай бұрын
1. vector database 2. embedding
@generativegeek Ай бұрын
you are correct. please share your email address by sending an email to [email protected]
@priyankasrivastava8356 Ай бұрын
hey can you provide that ppt 😛
@generativegeek Ай бұрын
request you to please take screenshots and build one yourself. i wont be able to share the ppt at this moment, sorry. :(
@priyankasrivastava8356 Ай бұрын
@@generativegeek no issues, bdw thanks its very helpful I am implementing all to understand more better way
@generativegeek Ай бұрын
please let me know if you have any doubts / questions. happy to help.
@priyankasrivastava8356 Ай бұрын
@@generativegeek It seems free services from open ai API is stopped am I right ? second one if I use google API instead of open ai what I need to change actually I implemented whole but when I used that last llm part its shows error
@generativegeek Ай бұрын
yes, you can however load $5 credits in your open ai account and it will last for a very long time (400 rs roughly). If you have the code / colab notebook / github repo please share and i will look into the error you might be facing. Alternatively mail it at [email protected].
@priyankasrivastava8356 Ай бұрын
in given link of note book it is not completed it seems
@generativegeek Ай бұрын
thank you for bringing it up. i will fix it in a couple of days. Seems like the notebook did not autosave :(
@priyankasrivastava8356 Ай бұрын
@@generativegeek ok thanks for responding
@priyankasrivastava8356 Ай бұрын
amazing i understand each and every thing in detailed view thankyou so much
@generativegeek Ай бұрын
thank you so much for your appreciation. i am glad you liked it.
@srdpt3 Ай бұрын
can you share the code?
@generativegeek Ай бұрын
here you go : github.com/technofriends/video-to-clips-langchain-demo
@weorganicsphere Ай бұрын
Can you pls share the code?
@generativegeek Ай бұрын
here you go : github.com/technofriends/video-to-clips-langchain-demo
@ralessungaminf8892 Ай бұрын
1. Embedding 2. Vector DB
@generativegeek Ай бұрын
thank you. you got the sequence wrong though :(
@vaibhavlehra1482 Ай бұрын
1 embedding 2 vector db
@generativegeek Ай бұрын
thank you. you got the sequence wrong :(
@avirajput26 Ай бұрын
1. Vector database 2. Embedding
@generativegeek Ай бұрын
you got it right. Please share your email address by dropping me a note at [email protected] and will send your voucher to you.
@jlfguthrie Ай бұрын
Thank you for a very good tutorial. Am I understanding correctly, if I select a vector store component, like Chroma, does Langflow automatically create the Chroma database, or do I have to have a seperate Chroma db elsewhere?
@generativegeek Ай бұрын
Thank you. langflow will persist the chroma db in the path you specify. You will be able to see it within the working directory. At 10:26 in the video you can see when i bring in the chroma langflow component there is an option to specify persist directory.
@manasnapte1 Ай бұрын
Thanks Vaibhav for a great video! I am currently working on a use case to synthesize Knowledge articles out of KZfaq videos. (Ex: I select 5 videos from KZfaq explaining a concept or elaborating on a topic and I should be able to directly get Knowledge articles from them as an output) After the transcript generation as in the video I will create a prompt to create summaries and knowledge articles directly. have you come across similar use case earlier?
@generativegeek Ай бұрын
Hi @manasapte1 thank you for your appreciation. Am glad you found the video useful. yes the use case you have mentioned is a valid use case and we have deployed something similar on production for our product. I will suggest you also go through this video kzfaq.info/get/bejne/bZZxYLeSlc2seZ8.html to understand how to do summarization of several documents.
@machanmaxx Ай бұрын
great explanation. keep it up ❤
@generativegeek Ай бұрын
thank you very much for your appreciation and kind words.
@WanderlustVibes777 Ай бұрын
How can we get hyperlink along with surrounding text and a table while extracting text from pdf using python if you could make a video on this
@generativegeek Ай бұрын
I will give it a try. However, have you tried PyMuPDF library? pymupdf.readthedocs.io/en/latest/ Also additionally look at github.com/jsvine/pdfplumber
@karthickdurai2157 Ай бұрын
great video
@generativegeek Ай бұрын
thank you. am glad you liked it.
@skilz8098 Ай бұрын
I'd rather write all of the C/C++, OpenGL, Vulkan Code knowing that I built the application. I guess it's the engineering personality, perspective that I have.
@generativegeek Ай бұрын
yes, its a lot about what works for one. Generative AI is in a way democratising the skills and power that was limited to a few some time ago. Be it writing great headlines, great catchy stories and content or just coding and building PRD's for product specs. In my personal view, the future is full of opportunities for those like you who are already significantly ahead in the curve.
@eternal_dread666 Ай бұрын
what is this supposed to do after?
@generativegeek Ай бұрын
if you have an application in mind that you want to build where your knowledge base will also have KZfaq content, then this can be a great way for you to bring that content into your vector / other kind of storage for efficient retrieval later.
@eternal_dread666 Ай бұрын
@@generativegeek can u make automated videos with js/python and autogpt?
@generativegeek Ай бұрын
can you specify some use case you want to see a video about?
@eternal_dread666 Ай бұрын
@@generativegeek someone automated his channel Using JavaScript/chatgpt/ffmpeg and others
@generativegeek Ай бұрын
got it. It can be done. I will try to build this and post a video on this soon. I am on a travel schedule right so give me 5-7 days.
@d3n4c3 Ай бұрын
Kind of crazy though, I had gpt-4o literally code me a fully functional Tetris game with the prompt: "Code me a Tetris game in python." It worked the first time, keeps score, has a game over text when you lose, and says press any key to play again.
@generativegeek Ай бұрын
yes its awesome what it can do with just a simple prompt. Try giving it a screenshot of some feature you want implemented and ask it to update its look and feel and code basis the screenshot. you will be amazed how much it can infer from such little context.
@explorewithrudransh8342 Ай бұрын
This is a great tutorial! I enjoyed making the game and learning the skill of making other games using ChatGPT. Thank you so much for the fun and educational videos. Looking forward to the next one!
@generativegeek Ай бұрын
thanks a ton for your appreciation. I am glad you liked the video and learnt something useful. god bless you.
@finnical9718 Ай бұрын
"Even a no-code developer could probably do this iteratively back and forth!" I am literally doing that. I'm not even using the paid version (yes it's a lot harder that way).
@generativegeek Ай бұрын
lovely. yes, thats the power of large language models. What other games or apps have you built?
@finnical9718 Ай бұрын
@@generativegeek Literally none. I am programming a board game I made in Unity, thanks to ChatGPT. I'm...not very far yet. I'm busy being a stay-at-home dad.
@generativegeek Ай бұрын
would love to hear more about it once you near completion. all the best.
@finnical9718 Ай бұрын
@@generativegeek Expect a long timeline! But thank you for your interest, I might actually get to the point of selling it one day.
@generativegeek Ай бұрын
all the best @finnical9718. Building something and then being able to sell it is the peak satisfaction moment for any builder.
@IlllIlllIlllIlll Ай бұрын
Can you make a video on this scenario i just invented: make a chatbot using either claude 3 ai / or chatgpt.(this is what i have) where you can find a way to train it based on youtube videos, documentations so it will respond accordingly to the information it is provided. Once you finish the API, you will provide it to your friend who created the chat app for ios and android but was just missing the API key. I can provide more information if i message you on linkedin. I havent learned to code yet but i want to see if you can teach me on developing the API. So once i create the chatbot i like. I just provide the api key to app developer so he can insert in app so users can use it to ask questions etc
@generativegeek Ай бұрын
yes this is possible. i will try to find time to build this video over the next week. However, if you don't know coding, it will be tough for you to make sense of it. This project that you have described is more involved than building simple screenshot derived games. Will ping you back here once the video is live.
@IlllIlllIlllIlll Ай бұрын
@@generativegeek can you try to use Claude 3 haiku and chatgpt to try to generate the code. I noticed chatgpt has a lot of different gpt's although maybe 4o is better than all those gpts available.
@generativegeek Ай бұрын
sure. let me try that out and post a video. It will be a nice comparative analysis.
@KA00_7 Ай бұрын
informative✌️
@generativegeek Ай бұрын
thank you for your appreciation.
@explorewithrudransh8342 2 ай бұрын
This was a great video! I really learnt a lot from the detailed explanations and understood more about the basics of GPT -4o. Thank you so much! It was really fun making this game!!
@generativegeek 2 ай бұрын
thanks so much for your encouraging comment. god bless you.
@thinker-pd6xh 2 ай бұрын
Sir, can you make a video on web scraping and integrating it with streamlit? For example: we should be able to select any reviews like google, amazon and flipkart. And the if we pass the url, it should web scrap and then extraxt the reviews and do sentiment classification(using gemini ai or anything) on the reviews and send it as output. The web scraping videos on youtube are diff to understand and would love to know if there are any simpler methods like Autoscraper(its not working properly on all urls). Thankyou. Looking forward to learn alot from you sir.
@generativegeek 2 ай бұрын
thank you and sure this is possible. I will prepare this content over the next week. Allow 5-10 days for this as there is a backlog of 3 other videos i am working on right now :).
@generativegeek 2 ай бұрын
in the meanwhile, can you see if this video helps your goal? It doesnt have the scraping part but everything else is there to help you build a solid foundation - kzfaq.info/get/bejne/gLt9a7B50bWYpGg.htmlsi=8npVu0lqQAvC21j_
@thinker-pd6xh 2 ай бұрын
Sure sir, will definitely check out. Thankyou.
@thinker-pd6xh 2 ай бұрын
Loved it! Thankyou sir. Your channel is awsm and beginner friendly.
@generativegeek 2 ай бұрын
thank you so much for your kind words. am glad you like the content.