I don't know why, but I find these videos relaxing
@CM-zl2jw
Ай бұрын
😂. Panic. Igor recently told me he is one with the fire hose. Are you feeling hosed yet? 😂
@juicegod777
29 күн бұрын
Same, just fell asleep watching
@02167631
11 сағат бұрын
With the sea of countless AI tools, so many of us are a) waiting to see who comes out on top b) suffering from confusion because of all the options available. Most of them being inferior! Watching your videos allowed me to be the hero to my young nephew. He needed pictures for some classwork and I suggested using KlingAI. It was not only better than the model he was using, he got an A on his assignment!
@justme-wf5hx
Ай бұрын
Thank you so much, great video and channel, happy to see you succeeding in what you're doing.
@martinpercy5908
Ай бұрын
Such a great video - well done Igor and team, commenting for the algorithm and joining the channel...
@aiadvantage
16 күн бұрын
Appreciate that!
@artifishially_stupid
Ай бұрын
Thanks Igor for a very understandable discussion on fine-tuning. I've been studying into this topic on and off for a while now, and you explained it better than anyone. Now that I see what it is, I'm a bit disappointed. I expect an LLM to be good enough to access my clean, preprocessed Knowledge Base files and use its "intelligence" to answer user questions SO I DON'T HAVE TO CREATE A GLORIFIED FAQ. Fine-tuning, in my opinion, is kind of like owning a Ferrari but instead of driving it around town, we're pushing it around town.
@jeffg4686
Ай бұрын
the fine tuning gpt4o is the winner this week oh, actually,maybe it has to compete with heygen's interactive avatars Regarding the videos, how about someone create an AI program to fix the issues in the generated content. post production AI - like audio cleanup, but video cleanup. it would add some frames where needed, and remove some existing ones, without outcome being a seamless video that actually works. Tell them to make that part of their video editor they are working on (fix the video first, then edit)
@CM-zl2jw
Ай бұрын
The loop feature is indeed very cool. When I come up for air I will have to check it out. 😅
@Sir.Black.
29 күн бұрын
The editing team showing off just to make sure they keep their jobs lol
@coreyostman
Ай бұрын
At the moment, you have to be API Tier 4 in order to fine-tune gpt-4o mini. Hopefully they'll roll it out to folks that haven't spent $250 on API yet.
@monomer888
Ай бұрын
Thanks for pointing this out! I got the error message about needing to be in Tier 4 when I tried to have a go.
@Ben_D.
Ай бұрын
for the 'rythm 👍
@timmcgirl5588
Ай бұрын
Open Interpreter + Llama 3.1 + n8n + Gorilla AI = 100% autonomous agent that automates all workflows, all open source and free, access to over 1600 API's.
@Niiwastaken
Ай бұрын
Tbh i dont consider the inference of local LLM's that slow and im on a 1070 ti. Its still MUCH faster then when talking to humans online
@PersianMate
Ай бұрын
Hi, I have a business-related question; how can I contact you?
@nftjakob5187
Ай бұрын
Good video, bro! I have a question. If I have to create a GPT that is an expert and really good at knowing/answering all the Gold rules, what is the best way? Is it fine-tuning the model or is there another way?
@user-qg8qc5qb9r
Ай бұрын
Introduction and Overview of AI News - 00:00:00 Llama 3.1 and Hugging Face Chat Interface - 00:00:30 ChatGPT-4 Mini Fine-Tuning Tutorial - 00:02:20 Understanding Fine-Tuning and its Benefits - 00:02:48 Practical Example of Fine-Tuning with JSON - 00:03:45 Using OpenAI's Interface for Fine-Tuning - 00:05:34 Llama 8B Fine-Tuning Possibilities - 00:07:26 SearchGPT Waitlist and Details - 00:07:50 Mistral Large 2 Model Release and Analysis - 00:08:45 Differences Between Open Source and Mistral License - 00:10:10 Benchmarks and Initial Impressions of Mistral - 00:11:06 Interactive Avatars from Haen Labs - 00:12:27 Suno's New AI Music Feature: Stems - 00:13:44 AI Video Generation Updates - 00:14:38 Kling AI Model Accessibility and Performance - 00:15:13 Comparison of Various AI Video Generators - 00:17:18 Luma Labs Update: Key Framing and Loops - 00:18:45 Demonstration of Luma Labs' New Features - 00:19:06 Photoshop and Illustrator AI Updates - 00:22:02 Bringing AI Innovations to Everyday Life - 00:22:53 Conclusion and Final Thoughts - 00:24:15
@Candyman2020_
Ай бұрын
@Renee_egan
Ай бұрын
OMG I need to get that fold phone! Was it Motorola??? Wanted one BEFORE they the ai was added to it, now I may have to hop ship from apple 😱💀 then again idk if adding only ai image rendering could make having one single device outside of the apple ecosystem worth the huge communication issue it would create, probably not… Anyone found ways to merge ai into Siri on iPhones yet? lol I'm back on break and just got a bit further into this video but guess hey artist one more back to finish this, brutal short breaks
@Renee_egan
Ай бұрын
Love the video and oh a new llama and Igor photo! 😂😂 I have a question (and possibly a tutorial idea) : I have a 2018 MacBook Pro that can no longer support Apple’s updates so I’m still running Ventura and I’m not able to download the app for GPT 4o is there a safe-ish way to do this myself? Also would love Suno power user best practices. Also how do you take suno output (whether stems, compete result from a promt etc) and if needed convey the format for use on mobile (such as to create a custom ringtone and custom alert/notification tones) how we can apply these tones to Alexa or google home (if that's possible?) I HATE those sleeves tones with a burning passion! theyre a guaranteed way to instantly turn me into a bitch soon as I woke up if that's what I wake up from lol on the flip side; it has seemed to subconsciously train my body into a morning person who wakes up before the alarm goes off 😂 yes I hate them that much
@nuhmelik9360
Ай бұрын
I can't access the llama, I live in Denmark but I can't access it. It's annoying. 😡😡😡
Пікірлер: 27