Learn AI Engineer Skills: Autonomous Agentic Behavior (Llama 3 8B Ollama)
HTML-код
- Опубликовано: 29 июн 2024
- Learn AI Engineering Skills: Autonomous Agentic Behavior (Llama 3 8B Ollama)
👊 Become a member and get access to GitHub and Code:
/ allaboutai
🤖 Great AI Engineer Course:
scrimba.com/learn/aiengineer?...
🔥Open GitHub Projects:
github.com/AllAboutAI-YT/easy...
📧 Join the newsletter:
www.allabtai.com/newsletter/
🌐 My website:
www.allabtai.com
In this video I talk about how to create a basic autonomous agentic system that using a LLM to solve simple task. The python system can be expaned to include more function and tools. Using llama 3 8b on ollama and Claude 3 Haiku
00:00 Autonomus Agentic System Intro
02:22 Explaining System / Code
16:49 Agentic System Test 1
19:05 Adding a Tool to the system
19:46 Agentic System Test 2
21:11 Conclusion Наука
You are a real life saver. I was trying to to figure out how to divide by elephants for my job. I might just get a raise for this! Thanks
🤣🤣
I love how the AI finds the average weight of an elephant. This LLM is just going to replace your job if you don't adapt...
Another great one. Keep em coming! You have interesting thought process, thank you
This is getting more and more my most liked channel. Keep it up !
Useful information. Thank you!👍👍👍
Solid model for these tasks.
How does the system handle invalid JSON responses during its operation?
Please make video about Romo AI super AI tools in one place and learnex AI world's First fully AI powered education platform. My favorite AI tools
Task for You: to get 2 chatbots (mistral and claude LLMs?) to discuss with each other , dialoguing and figuring out with each other's arguments and reasonings, and convincing each other on the 5 most important Human projects that could change Humanity to the better in the next 2 weeks....
Let 2 Ai Chatbots figure this out with each other !
i wonder what thes 5 resulting projects would become
thanks!
Ok.. new kid on the block here... where do I find this example on the members github?
Hi Kris, paid for a Master (14,99) sub but didn't recieve anything...do you need to aprove or something?
Hi Kris. I have heard that there are some versions of llama3 that have a very large context window. If this is true, is there an easy way to feed a bunch of articles into the model using RAG and query it easily? An example implementation might be a troubleshooting system in which there might be a knowledge base with many articles on common troubleshooting steps for varied issues where the KB is all read into the LLM and someone can ask questions of the LLM and it can determine the most likely cause based on the problem description. Any suggestions?
higher context requires more ram so keep that in mind just search on hugging face for llama3 gguf models
I think Kris has a video on this. You have to generate an optimized query (by AI), then feed your docs chunk by chunk together with your query/question piece by piece into the context. Each step creates a summary which is then analyzed in the end.
Still 10000% useless compared to the future i came from
You forgot to divide your percentage by elephant weight. Repeat query.
I know, it's like living in the Dark Ages before we had replicators and point to point wormholes