I have spent over 3000 hours learning and coding over 300 projects and share everything I have learned in these KZread 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
Пікірлер
Patreon my *ss!
How in the world do you get something out so quickly, every time?
Experience and cursor. Superpowers are real :) thank you 🙏
Learn to code fast with my 1000x MasterClass:www.patreon.com/posts/1000x-dev-23rd-3-103326330 Talk with me this Sunday the 28th, AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meetings-for-108628691 Search 200+ echohive videos and code download lnks:www.echohive.live/ Chat with us on Discord: discord.gg/PPxTP3Cs3G Follow on twitter(X) : twitter.com/hive_echo
as awesome as it is simple, love it!
Thank you very much 🙏
Download the full Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/self-instructing-108628219 Talk with me this Sunday the 21th, AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meetings-for-108628691 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Search 200+ echohive videos and code download lnks: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
It's a pleasure to get to join this stream of consciousness as you build new stuff almost every day. Thanks for sharing everything that you do! 🙏🏾💜
Thank you very much. It is a pleasure for me as well :)
Download the full Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/exa-web-with-in-108466079 Talk with me this Sunday the 21th, AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meeting-for-107298815 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Search 200+ echohive videos and code download lnks: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
Can you please make a video about permission based deepfake models ?
I replaced Claude Sonnet 3.5 by GPT-4o mini api for coding and it was a bad idea, GPT-4o mini is bad like really bad. Don't trust this model if you don't waste your time, stick with the best model for coding at least.
Yeah it was my experience as well. This model is useful for batches of non critical language tasks like structured data etc especially at this cost level.
New here and this is some mind blowing content! Is there a course or other videos about how this agents are being built?
Thank you 🙏 take a look at a variation of this: 10+ Agents work together to build any Python app kzread.info/dash/bejne/qZ6i1NCNdMzLmNI.html
@@echohive Thank you! I will check it out
Thank you sir for the great content. Do the iterations act as a control as it is possible to run an endless loop?
Surely you could but that doesn’t mean it produces something better.
hey this is useful,can you share the git repo link?
can this be run all locally at all? meaning all the Hume interactions only run on local servers and have no web access?
No, unfortunately this is currently only possible via Hume api. But I am sure in the future there will be loca open source alternatives.
Thank you, this is another great video this new mini model, just released really seems to perform well at a fraction of the cost. I also loved the projects and how receipts can be uploaded and automated into a database. Excellent content as always!
Thank you 🙏 the price point is also amazing
it looks as a useful tool for data extraction from long documents and change in to JSON format - for resoning I will still use Claude 3.5 :)
You are spot on! For data extraction, it is very good although not perfect and super cheap 🙌 plus 16k output
Download the full Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/gpt-4o-mini-data-108414519 Download Multi Modal Image Search files as a free member at my Patreon: www.patreon.com/posts/code-files-for-97340865 Talk with me this Sunday the 21th, AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meetings-for-108258004 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Search 200+ echohive videos and code download lnks: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
It's funny that before computers came along we had committees. These were used to validate stupid decision based on the assumption if a group makes a decision individuals can't be held accountable. On top of that, it was recognised a long time ago that a decision reached by committee was likely to be bad compared to a decision made by a single expert on the subject. Now we have hallucinating, misinformed and massively inefficient AI failing to deliver so the solution is to create a committee of AI. Maybe that works for fiction etc., but for everything else? Let's see...
Can you create a video about llamafile new open source project? kzread.info/dash/bejne/X6GGy497Y9eclqQ.htmlsi=7X43cc_zxIXyFmlO
Great to see the reasoning steps!
Nice:)
Thank you 🙏
Download AGENT AGI at my Patreon along with 250+ other projects:www.patreon.com/posts/agent-agi-builds-107481847 Talk with me this Sunday the 21th, AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meetings-for-108258004 Download a basic version of Agent Agi: www.patreon.com/posts/agent-agi-no-no-107929269 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Other Agent AGI videos: kzread.info/dash/bejne/k6SMucODZ9XbqbQ.html kzread.info/dash/bejne/mJ2H0rSCmtXUerA.html kzread.info/dash/bejne/npOBzcGjiZa6pqQ.html Search 200+ echohive videos and code download lnks: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
aight thats enough. for the first time in my life, I'm joining a f*kn patreon.
Thank you for your support if you do 🙏 I have actually made improvements on Agent AGI since this video. kzread.info/dash/bejne/mJ2H0rSCmtXUerA.htmlsi=r-q4_f8njz_vjsVo and I try to update it regularly on Patreon more often.
i dont have money to get access
Thanks Sir, Sir so many of us cant afford for code pricing anyway thanks for nice content
Thank you 🙏 feel free to send a message to me on discord.
Is there a way to get the token cost down during testing and development?
You could try to use smaller models like haiku, Gemini flash etc. but then the output will be of lesser quality.
Im so happy to see you continuing the multi-agent coder series!!! This last one is perfect. Sets the foundation for lots of config with agent ensemble 💜 This is exactly what i needed to lift my spirits after a rough couple weeks
Thank you 🙏 and I am sorry to hear that you had a rough couple of weeks. This one is a fun one. Let me know what you think when you try it
@@echohivewill do 😁
@echohive, how do we communicate/ask question with you? I signed up with Patron but send direct message seams to be disabled, so i asked a question on the community but the last time you were there was over 1 month ago. What would be great is to concentrate on a single project like this video above and constantly improve based on Parton feedback.
Thank you for being a Patron 🙏 You should be able to send me a DM at Patreon. I get messages from other patrons. Sorry about the community chat. I don't check there often as it is not very busy there. But you can join the discord. I check messages there every day discord.gg/7BNTCsyt. take a look at this video which has a automatic improvemtn cycle: discord.gg/7BNTCsyt But this particular project also has a feedback cycle after the error correction phase
Good. However, the glaring issue with AI coding is not the creation of code from scratch. Large Language Models (LLMs) can handle that task quite effectively. Instead, I believe we should focus on improving existing code, which requires a comprehensive understanding of the project's context. What do you think?
yeah, that is a good point. This project can pick up from a previous file as well. Also check out the agentic-dev-team video as well which handles multi file python projects. I will continue to explore and refine the process
A paper worth checking out. Agentless: Demystifying LLM software Engineering Agents.
Thank you for sharing 🙏
Download the Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/gemini-1-5-pro-3-108226932 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Search 200+ echohive videos and code download lnks: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
sickkkk
Thank you 🙏
That's so strange that they seem to have a bias for themselves
Yeah that was a bit bizarre haha. They must know their own tokens 😜
This is so cool!! Very useful for setting up multiagent stuff. I'm glad we're able to name the models in this implementation
Thank you 🙏 let me know what you think when you try.
Great - ) I wonder it is possible to build multimadal worflow - for example Claude 3.5 (content creator base on provided context) GPT4o (content veryfication, critics idea for improvment), Geminy 1.5 Pro (final content transaltion) :)
Thank you 🙏 yeah I will think of a way to add some nice multimodal support. Maybe with a window as I did with agent AGI
🎉🎉🎉litumezi (thank you).
Thank you too :) 🙏
🎉
Thank you 🙏
Amazing job
Thank you very much! 🙏
Download the full Project files for this project at my Patreon along with 250+ other projects:www.patreon.com/posts/unified-llm-apis-108090093 Download the basic Project files without examples for this project at my Patreon along with 250+ other projects: www.patreon.com/posts/unified-llm-apis-108094414 Talk with me this Sunday the 14th AMA for Architect+ level Patrons: www.patreon.com/posts/ama-meeting-for-107298815 Learn to code fast with AI assistance with my 1000x MasterClass: www.patreon.com/posts/1000x-dev-103326330\ Search 200+ echohive videos and code download lnks: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
Just wonderful. I have a book I'm writing and I'm 20% into it. So I can use the content from what I've written as the grounding text?
Thank you 🙏 Yeah you should be able to but you would want to change the system message as it is instructed to create new characters currently.
Next, agent animation maker, then agent movie maker, and finally agent VR and AR video game maker. 😎🤖
Haha I hope so :) agent systems are really fun to make.
ONE SENTENCE SUMMARY: This video showcases the capabilities of the agent Dave team, a system involving multiple static agents and dynamic processes to build applications like calendar, chat GPT clone, and diagram tools through AI planning and error correction. MAIN POINTS: 1. The agent Dave team consists of four static agents with no limit on dynamic agent creation. 2. Static agents are divided for planning stages, generating the project plan including file needs. 3. Dynamic agents write individual files in parallel without knowing others' work. 4. An error correcting agent fixes identified errors by revising and testing code until all issues resolved. 5. A feedback loop adjusts application based on user interaction and provided inputs. 6. Over 900 lines of code generated, demonstrating complexity handled by the system. 7. The project requires adherence to guidelines for file naming conventions and main function usage. 8. Access to full code is crucial to facilitate understanding and further development. 9. Becoming a patron grants exclusive access to detailed code files and courses on AI programming. 10. Streamlit course, Fast API course, THX Master Class are some of the learning resources provided. TAKEAWAYS: 1. The agent Dave system demonstrates AI's capability in developing multifile applications through specialized planning and correction phases. 2. Patrons gain access to comprehensive code repositories for inspiration and hands-on coding experience. 3. The system uses anthropic API for interaction, highlighting potential risks while showcasing innovative AI usage. 4. Attention is directed towards specific file management practices such as naming conventions and main function adherence for successful project creation. 5. The THX Master Class offers insights into efficient programming techniques. 6. The error correction phase improves application quality by automatically addressing issues before user testing. 7. Patrons can connect with creators for one-on-one consultations, enhancing the learning experience.