LLAMA 3 Released - All You Need to Know
Вставка
- Опубліковано 31 тра 2024
- MetaAI just released LLAMA3, the best in class LLM for its size. Here is the first look
🦾 Discord: / discord
☕ Buy me a Coffee: ko-fi.com/promptengineering
|🔴 Patreon: / promptengineering
💼Consulting: calendly.com/engineerprompt/c...
📧 Business Contact: engineerprompt@gmail.com
Become Member: tinyurl.com/y5h28s6h
💻 Pre-configured localGPT VM: bit.ly/localGPT (use Code: PromptEngineering for 50% off).
Signup for Advanced RAG:
tally.so/r/3y9bb0
LINKS:
Announcement: llama.meta.com/llama3/
Meta Platform: meta.ai
TIMESTAMPS:
[00:00] Introducing Llama 3: Meta's Latest AI Model
[01:28] Training Data and Context Length Improvements
[03:58] Technical Insights and Human Evaluation
[04:42] Future Prospects for Llama 3
[05:42] Testing the model
All Interesting Videos:
Everything LangChain: • LangChain
Everything LLM: • Large Language Models
Everything Midjourney: • MidJourney Tutorials
AI Image Generation: • AI Image Generation Tu... - Наука та технологія
Finally, our requests were heard! The models seem pretty decent. Can’t wait to check them out for myself. And I just hope that with a little patience we’ll eventually get more context length
Dude there was once a time I used to get news from Google about the latest tech updates. But now you have replaced Google for me 🤣
Great work 💯
haha, that's the best compliment I have gotten :)
Kudos to the great work you are doing ! Big fan of localGPT !
Thank you 😊
On Reddit say, Ollama have Llama 3 ready to download and run
Great Video, thank for sharing
Thanks 😊
Impressive for "Sally" example 🤩! Haven't seen this in other open models, right? Have u seen? Other models answer 2, then when u bring up the objection they understand they can't take it from context.
Yup, it's smart, I tested wizradlm (last video on the channels), and had to remind it
@@engineerprompt and less verbose in comparison with wizardml. I saw ur other video and tweeted about it. Crazy time.
Im waiting for 8x70b Mixture Of Experts
Actually its interesting that Meta didn't go for a MoE, even the 400B version seems to be a dense model not an MoE.
What's about new Stable Diffusion 3? Is it connected to Llama 3?
there is an image generation model on meta.ai, figured that out later but not sure if its based on stable diffusion.
I've found the 8B model is not censored against ERP
are you controlling via system message? I heard the same and want to test it out.
Thank you for your video. Could you add your script for forigners?
Thank you, good idea. Will start doing that
Waiting for local gpt integration of llama3 8b
that's coming soon. I will have a busy weekend :)
What is Llama 3?
Llama 3 is a large language model which is trained on large amount of data on many GPUs and training take weeks and months and right now it is best open source large language model
8k context window. not impressed.
also I think they've compared the 8B with mistral 7B o.1 not the new o.2.
I agree but its really impressive that you can train an 8B model for 15T tokens. The scaling laws go brrr.......
@@engineerprompt 🤣 that's true
1