- 332
- 526 482
echohive
Registrace 6. 08. 2022
I have spent over 3000 hours learning and coding over 300 projects and share everything I have learned in these CZcams videos. I hope you will find them useful :)
search all echohive videos: www.echohive.live/
You can download the souce code to all my projects by becoming a Patron
search all echohive videos: www.echohive.live/
You can download the souce code to all my projects by becoming a Patron
Modular Rag and chat implementation from URLs, PDFs and txt files.
This project has many modular pure python functions to split text with overlap, embed text and cosine similarity and more so you can experiment and try to come up with your own rag strategies.
Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/modular-rag-and-106461497
Learn to code fast with my 1000x MasterClass:
www.patreon.com/posts/1000x-dev-103326330
Search 200+ echohive videos and code download links:www.echohive.live/
Auto Streamer: www.autostreamer.live/
Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684
Chat with us on Discord: discord.gg/PPxTP3Cs3G
Follow on twitter(X) : hive_echo
CHAPTERS:
00:00 Demo
05:52 My Patreon benefits
06:26 Code review
#rag #vectordb #ragchat #retrievalaugmentedgeneration #openaiapi #pythonai #gpt4o #artificialintelligence #largelanguagemodels
Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/modular-rag-and-106461497
Learn to code fast with my 1000x MasterClass:
www.patreon.com/posts/1000x-dev-103326330
Search 200+ echohive videos and code download links:www.echohive.live/
Auto Streamer: www.autostreamer.live/
Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684
Chat with us on Discord: discord.gg/PPxTP3Cs3G
Follow on twitter(X) : hive_echo
CHAPTERS:
00:00 Demo
05:52 My Patreon benefits
06:26 Code review
#rag #vectordb #ragchat #retrievalaugmentedgeneration #openaiapi #pythonai #gpt4o #artificialintelligence #largelanguagemodels
zhlédnutí: 575
Video
Cursor IDE deep dive and why it is the best tool for coding
zhlédnutí 2,4KPřed 4 hodinami
we will explore many of the awesome features of cursor, such as cursor chat, ctrl k command, long context and interpreter modes, as well as documentation vectorizer, definitions command and .cursorrules file Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Download the Project files for this project at my Patreon along with 250 other projects:www.patreon.c...
Multiple LLM Agents perform analysis on an ongoing conversation. Next JS project with Vercel AI SDK
zhlédnutí 1,8KPřed 12 hodinami
In this Next JS project we can have up to 8 agents each with a custom system message which can perform any type of analysis on the ongoing conversation based on default system messages or from your custom instructions. Download the Project files for this project at my Patreon along with 250 other projects:www.patreon.com/posts/multiple-llm-on-106220356 Learn to code fast with my 1000x MasterCla...
Easy web Scraping, crawling, web search for LLMs with GPT-4o chat
zhlédnutí 755Před 19 hodinami
Easy web Scraping Crawl Search with Firecrawl including chat with GPT-4o using the context ALL Project files for this project at my Patreon along with 250 other projects:www.patreon.com/posts/all-files-for-4o-106035051 BASIC Project files for this project at my Patreon along with 250 other www.patreon.com/posts/basic-files-for-106035049 Learn to code fast with my 1000x MasterClass: www.patreon....
Easy Web scraping + search with Jina AI extracts clean text for LLM context including PDF urls
zhlédnutí 879Před 21 hodinou
Jina AI which you can use with or without an API key which returns a clean text from urls or web search. Project files for this project at my Patreon along with 250 other projects www.patreon.com/posts/easy-free-web-ai-105970057 Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/ Auto Strea...
Open Source Multi Modal Rag with chromadb using OpenCLIP embeddings
zhlédnutí 704Před dnem
Multi Modal Rag system with chromadb using only opensource embeddings Code files and coding walkthrough for this project available at my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: di...
GPT-4 Omni personas perform market research for you with real-time updating plot and coding walkthru
zhlédnutí 730Před dnem
In this script we create "n" many personas based on a specific demographics and then ask these personas a particular question with given options. Then we plot them in real time and save the results to a json file Learn to code fast with my 1000x MasterClass(code included for this project): www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive....
My personal strategy for coding to prepare for GPT-5.
zhlédnutí 1,1KPřed dnem
My personal strategy to prepare for GPT-5 NOTE: list of example things to learn is meant to only serve as an example Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: dis...
Livekit real-time Ai agent voice interface in Python with Auto detected user interruptions
zhlédnutí 1,6KPřed dnem
Livekit real-time Ai agent voice interface in Python with Auto detected user interruptions Download Project files for this project at my Patreon: www.patreon.com/posts/livekit-real-ai-105674470 Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course:...
Next js + FastAPI + Python implementation of realtime chat sentiment analysis with GPT-4 Omni
zhlédnutí 723Před 14 dny
This video features real time chat sentiment analysis in 3 implementations. Python terminal, Fastapi webapp and Next JS webapp Download all Project files for this project including Next JS app at my Patreon: www.patreon.com/posts/all-files-for-js-105610496 Download python and fastapi Project files at my Patreon: www.patreon.com/posts/python-and-app-4-105610500 Learn to code fast 1000x MasterCla...
Should you learn Next JS for LLM applications?
zhlédnutí 506Před 14 dny
Should you learn Next JS to build AI powered web applications? Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twit...
Gpt-4o chat app built with Vercel AI SDK + Next JS with streaming responses
zhlédnutí 774Před 14 dny
In this video we will built from scratch a chat app with streaming responses using the new Vercel AI SDK in Next JS Download Project files for this project at my Patreon along with 250 other projects www.patreon.com/posts/gpt-4o-chat-app-105468542 Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200 echohive videos and code download links:www.echohive.live/...
How does Phi-3 mini, medium and Gemini 1.5 Pro and Flash stack up in reasoning?
zhlédnutí 339Před 14 dny
How does Phi-3 mini, medium and Gemini 1.5 Pro and Flash stack up in reasoning? Project files for this project at my Patreon along with 250 other projects www.patreon.com/posts/how-does-phi-3-1-105341908 Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 previous reasoning test videos: czcams.com/video/_0enge5xpqA/video.html czcams.com/video/utodsdA_au8/video.html c...
Creating a Dynamic Knowledge Distillation Web App Using FastAPI and GPT-4o
zhlédnutí 547Před 14 dny
This is a fastapi webapp implementation of my previous knowledge distillation video which generates entire articles and markdown from gpt-4o generated sentences guided by the user given any topic Project files for this project at my Patreon along with 250 other projects www.patreon.com/posts/fastapi-webapp-105261864 Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330...
GPT-4o builds articles and markdown by first distilling knowledge sentence by sentence
zhlédnutí 1,3KPřed 21 dnem
In this script, user enters a topic then GPT-4 generates sentences. User can then expand on any sentence continuously. then user can decide to export the sentences as a text or markdown article Project files for this project at my Patreon along with 250 other projects www.patreon.com/posts/gpt-4o-builds-by-105177598 Learn to code fast 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330...
GPT-4 Omni creates DEEPER Knowledge trees based on any topic
zhlédnutí 493Před 21 dnem
GPT-4 Omni creates DEEPER Knowledge trees based on any topic
GPT-4 Omni generates unique and detailed chat personas. includes saving and retrieving of personas
zhlédnutí 579Před 21 dnem
GPT-4 Omni generates unique and detailed chat personas. includes saving and retrieving of personas
GPT-4o auto generates html landing pages from Streamlit webapp
zhlédnutí 942Před 21 dnem
GPT-4o auto generates html landing pages from Streamlit webapp
GPT-4o creates Knowledge trees based on any topic
zhlédnutí 2,9KPřed 28 dny
GPT-4o creates Knowledge trees based on any topic
TOP 15%!! GPT-4 Omni solves Space ship Titanic Kaggle data challenge
zhlédnutí 709Před měsícem
TOP 15%!! GPT-4 Omni solves Space ship Titanic Kaggle data challenge
Can GPT-4o watch and understand 10 videos at the same time?
zhlédnutí 1,2KPřed měsícem
Can GPT-4o watch and understand 10 videos at the same time?
Swarm Intelligence with GPT-4o, Gemini 1.5, Opus and Llama3 70b
zhlédnutí 1,6KPřed měsícem
Swarm Intelligence with GPT-4o, Gemini 1.5, Opus and Llama3 70b
GPT-4 Omni Auto Coder instructed by GPT-4 omni
zhlédnutí 2,8KPřed měsícem
GPT-4 Omni Auto Coder instructed by GPT-4 omni
GPT-4o vs ALL in reasoning battle. Who will win?
zhlédnutí 863Před měsícem
GPT-4o vs ALL in reasoning battle. Who will win?
Auto Streamer FastAPI Webapp teaches any topic
zhlédnutí 1KPřed měsícem
Auto Streamer FastAPI Webapp teaches any topic
How to build multi threaded apps in streamlit with GPT API
zhlédnutí 625Před měsícem
How to build multi threaded apps in streamlit with GPT API
Testing good-gpt-2 and also-good-gpt2 in chatbot arena
zhlédnutí 2,8KPřed měsícem
Testing good-gpt-2 and also-good-gpt2 in chatbot arena
How to use Interpreter Mode and Terminal Commands in Cursor IDE so you can code FASTER!
zhlédnutí 686Před měsícem
How to use Interpreter Mode and Terminal Commands in Cursor IDE so you can code FASTER!
EMERGENCY video: i-am-also-a-good-gpt2-chatbot
zhlédnutí 4KPřed měsícem
EMERGENCY video: i-am-also-a-good-gpt2-chatbot
Oooo I'm so excited for where you're taking this project! Am traveling right now. Will defo pick up and play with this once back home.
Thank you 🙏 have safe travels!
Have you looked into using postgresql with the pgvector extension? It allows using several embedding models, including openai. I've navigated away from storing things in Jason files or anything that is a file on my device and instead I embed everything in the actual database including images and base 64 and all kinds of metadata related to context. I'm going to be implementing PG vector to be able to take all context window information I've ever had for a user and creating embeddings that then allow for indexing and searching. And if you have not played with postgresql much, Open AI's model at least seems pretty knowledgeable about the use along with alembic and SQalchemy
I have used SQLite before but not Postgres with pgvector. But for quick projects using sql makes the code bulky and and queries are flat out confusing. Jason is a quick way to build a working thing then for production of course sql is the way to go or mongodb. Will look into pgvector tho. Thanks for mentioning it. Also lately I am really liking g chromadb. I hope they will get their cloud service up and running soon.
thank. you for the project:)
Thank you as well! 🙏
Amazing work echo. I am very glad that youtube has recommended your video to me, I have lost the link to your channel, but now I have subscribed. Looking forward to seeing you prgoress even more.
Thank you so much for the kind words and the encouragement!🙏
@@echohive You’re most welcome.
Oh Nice ❗❗❗ Have you played with Autogrok yet❓ I'd love to see an overview and use cases 👍
Thank you 🙏 I think I heard of it but didn’t try. Do you know the link to it?
Thank you!! You Are awesome!
Thank you too 🙏 glad you like the project
Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/modular-rag-and-106461497 Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/cursor-ide-deep-106337659 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
video on setting up interpreter mode would be useful. having trouble and so are others in the cursor github issue queue
Thank you for the suggestion. I will try to do that. Maybe with a post. But from what I remember you need to install Jupyter notebooks” extension. Then when you first run it will ask you to install “kernel” stuff and then it should work.
does it have any limitation for file sizes and token input lenght?
Yeah but I forgot the file suede Limit. Please cuffs the docs for that but it was pretty high I think like 200mb or something. Input token context is 128k with gpt-4o
more like this pretty please - great content
Thank you 🙏 any suggestions?
@@echohive maybe a deeper dive into some of those more advanced tricks. eg I had no idea about cursor rules & been using it for 6months! I only use it for js/ts files though, so guessing that's why I can't see interpreter mode... (but I also can't see the long context option) Cursor's docs aren't great... so these detailed examples are very helpful. Maybe also one on how you use cursor & git for version control? (maybe that's 100% obvious to you, but not us n00bs)
Thank you for the suggestions. I will look into making more videos with cursor. You can also check for an update to your cursor with ctrl+shift+p then search for update. Maybe your cursor isn’t updated to latest might be the reason you don’t see long context chat.
@@echohive all up to date - found it in the beta section of settings :)
Oh nice!
I’m using Cursor with GPT4o and Opus to code a major user platform in React. I’ve never coded anything more than frontend html and css. Using Cursor with the AI is surreal. It’s like having a code magician sitting with me. Like having 5 stackoverflow experts to ask any questions about any aspect of the project. I’m pretty sure now that I could build anything with it.
That is my feeling as well! Wish you the best with the project
cursor ide is great i have to say - exa search though appears poor, returning really weird results, lots of other services to use in this realm though- but good luck to them
Yeah, I do wish the search results were better :(
Great video! Do you personally use cursor pro or the API's?
Thank you 🙏 I have pro and also use AnthropicAI api key for opus as I found that is cheaper than the 10 cents per call after the 10 free calls per day.
@@echohive oh makes sense for sure. Do you think Opus is better than 4o for coding tasks?
Yeah usually it is better but slow too.
Have you heard of or tried continue dev. It's free and allows you to use open source models or just API calls.
Yeah I heard of it but didn’t try it yet
Cursor is my favorite AI assistant :)
Mine too! 🙂
You absolutely rock. Another great video . I already use cursor but I didn’t know it could do half of these features!!
Thank you 🙏 yeah Cursor has lots of open ended features which can be used in creative ways. I hope I was able to highlight them in this video. Happy to hear you enjoyed the video and glad you are using cursor. Happy coding! 💪
You took the words out of my mouth @rokvik. 20:01 This is incredible. Thank you, CZcams algorithm for finding this gentleman. What echohive’s teaching makes all the other channels in content I have found 20x more productive. 100x on his Patreon
Thank you so much! You are kind 🙂
Very cool, and great demo of v0
Thank you 🙏
Cool. I think you could also implement another layer of agents, which can Analyse all of the original Chat and the other agents in the layer below. One could make even more Abstraktion layers and distill a final answer to replace the original Chat answer.
Thank you. And yes! I do want to make some more experiments with setups like these and deeper ones too.
This is terrific !!!
Thank you so much! 🙏 glad to hear that :)
@@echohive You can! I am a computer scientist with 30Y+ experience, I wrote my first AI prog in 89. This tuto gave me a great idea for a quite complex project about mental health. To have it better understood by mgt I will display live REACT of different agents in separate windows. Thanks again. By the way I am learning next.js +Vercel, you got me convinced 😊
Thank you 🙏 I am happy you find an inspiration from this project. I resisted next js for a bit but at the end I had to try it and so far I really love it.
Seems to be just running Autogen in the backend using vercel ai sdk
Actually each agent is a straight forward api call to OpenAI using Vercel ai sdk. No autogen in this project
@@echohive ok.. can we use this setup and vercel SDK with react native?
I am not so sure about that. Please refer to Vercel ai sdk documentation. I am still learning next js and react and I don’t know of all the nuanced differences very well yet.
Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/multiple-llm-on-106220356 Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
I think it has same features with scrapingfish right?
I am not sure. I never used scrapingfish
Looks like this was a lot of work! Hope the channel is going well
Thank you very much! Yeah it was actually :)
Thank you:)
Thank you as well
ALL Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/all-files-for-4o-106035051 BASIC Project files for this project at my Patreon along with 250+ other www.patreon.com/posts/basic-files-for-106035049 Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
Great:)
Thank you 🙏
Omg this is super useful. Exactly what I needed TODAY 🙌🏾🔥
Yeah it is super easy and free too!
Project files for this project at my Patreon along with 250+ other projects www.patreon.com/posts/easy-free-web-ai-105970057 Learn to code fast with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
thanks for this test, since ur not even comparing 7bx8 im guessing its not even close to llama3, testing a q2 llama3 now as i only have 24gb vram
Great project - I will test it:)
Thank you 🙏 feel free to let me know what you think.
Oooo this is so cool!! 🔥 I'll definitely use this template for market research simulations
Thank you 🙏 let me know what you think
👏👏
Thank you 🙏
Is there a way to give it camera access so we can take a picture and load it as well?
Yeah that should be possible. Like taking a snapshot and storing it to the documents folder and rerunning the script. Or you can let is check the documents folder frequently during the script or before each query.
@@echohive I have a streamlit app that uses gpt-4o and has a camera feature that allows me to take a snapshot but you have to inquire about the image at once and it's not saved in memeory. This works better bc I can store multiple images before inquiring and they're saved to a database. I think checking frequently would be a little over-kill but works fine for your reasearch papers app where it's dynamically updated. I believe it works there bc papers are upload sporadically but with this, you actually know when data is laoded. So, if it had a UI I would just collect the images/data to the documents folder and have an update button which will rerun the script in the background. I travel a lot so i take a ton of pictures from the uber/lyft and it would be great to be able to find out more about them without Googling. I can think of tons of use cases.
Code files and coding walkthrough for this project available at my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
This video provided me with so much info thank you!
Very happy to hear that! Remember the list I made is not that important and you can and should make your own list of things to learn. Best of coding and learning to you 🍀
Do you think that frameworks like this, Vapi, Bland etc.. could get disrupted by openai's 4o api once it's multimodal/voice is fully released?
Yeah, I think so, but you never know. Good to keep an eye on this one and others
Next.js not bad but prefer „remix run”these days as js react framework….
will take a look at that one also.
What can it be used for❓❓❓ It seems answers are based on randomness or Bias. Sorry, I only watched a 2 minutes when i realized it wasn't based on reality, or polls, or real demographic numbers.
I use this type of stuff sometimes to decide on CZcams titles for example. It can be used in a variety of ways. I have a Webapp you can test which uses gpt 3.5 www.askaipeople.com
The more important part of the video is the live coding part of it In my mind.
So extremely apple bias on training?
There is new kind of marketing: you pay OpenAI to make gpt recommend your products😅 Also Apple is objectively a best option, so it could be just that
Learn to code fast with my 1000x MasterClass(code included for this project):: www.patreon.com/posts/1000x-dev-103326330 Search 200+ echohive videos and code download links:www.echohive.live/ Auto Streamer: www.autostreamer.live/ Fastapi course: www.patreon.com/posts/learn-fastapi-26-95041684 Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
This is legitimately great content.
Thank you very much! 🙏
Thanks for the info, couldnt fimd any documentation on the interpreter mode. Do you need to @codebase in interpreter mode for it to read it?
Ohh, I didn’t know we can use @ commands with interpreter. Yes actually, I just tested it and it works! You can also refer to individual files. Thank you for pointing this out! 🙏❤️
I finally just started really getting cursor going this past month and am really enjoying it! I really appreciate everything you have taught me this past year! I went from no code to making programs in multiple languages, formats, frameworks, making Intel Macs run GPU accelerators in C, Rust and Haskell, and building my own GPU server to host open source and soon my own models to run agent swarms. Really close to having my own version Hal 9000 at my house to do anything I want it to do. Echohive is one of the most forward thinkers on the web for AI implementation.
Thank you so much for the kind words! They made my day! 🙏 I am happy to be useful and I am glad you continued your learning journey and able to do all those things now 👏
Hi @echohive, thanks for the video. I'm looking for a tutorial on fine-tuning Mistral v3 or LlAMA 3 for NER using a JSONL training dataset. Does your Patreon have such an example/tutorial?
I have many examples with gpt-3.5 but none currently with other models. Will try to cover that soon.