Видео

Great vidéo ! thank you

Lots of buzz around Snowflake's "Arctic" LLM for Enterprises because it's Apache 2 licenses on Hugging Face. In theory it should work with LM Studio, but I wonder if anyone has actually tried it. 90% of what we're all looking to do is support Enterprise data, so this might be a powerful combination.

Its not al that good iv tested it

what is use of these...if we cant deploy?

Hey make a video that will make prompt to audio podcast using gpt-4o

its not crewai?

Is it possible to use praison AI with Autogen, I'm still on the waiting list for CrewAI

vectorising the document(s) and generating the response based on the prompt takes time for me, either using Chroma or FAISS, but yours went fast, any walkaround to ensure more efficiency and less runtime?

I create a video that tells people how to mind map with GPT-4o

1st one here...❤

This is scary. It could be turned into the ultimate surveillance tool.

…Germany??? …Happy??? 😂😂😂

is the API free ?

Assume audio only didn't find anything as the audio only doesn't have any input that tells it that the voice is that of Sam Altman so it's looking for a reference to him in the content

6:47 audio transcription with Whisper-1 model ???? Then GPT-4o does not process audio ???? After all, it was all supposed to be implemented in one model, in one neural network. Is it the TTS engine that is a separate model too?

Thanks mervin, am glad to have moved with you in Ai till this point, your tutorials have made me a better Ai user

See, this is what I meant. While we play around with, testing and leaderboards and oo-ah videos, OpenAI is build stuff and know we have watch more videos about what they built instead of building it ourselves. Everything was there, Groq, Whisper, gpt vision, etc. you want to make a lot of videos about nothing, go ahead.

Wow, Mervin you are amazing. Your videos are truly a value addition 🎉🎉🎉🎉

Thank you for your channel. I have a question for you. Local models are very powerful and necessary for privacy on local machines. However, if I wanted to expose the model’s endpoint via a Streamlit URL, how would you approach this project? Since the local model cannot directly communicate with a deployed Streamlit app, I am considering hosting the model from my local machine and providing a tunneling solution. While Ngrok is not suitable for commercial use without a plan, I am exploring alternatives such as Tor. 😊

Thx Mervin, would be nice to see the cost of the tests done.

Thanks for the quick overview, it was easy to follow. Can you give an idea of how long it took to process the video and audio? We couldn't tell from the edited version.

How can we output TTS audio with emotions like they shown in the GPT-4o keynote? When I look at the OAI docs it only show the old TTS model.

Marvin ia Chuck Norris of LLMs

Man you are quick Thank you for all of the content you put out..🙏

Input tokens will explode? Yes, great. Yes, why paying plus anymore…

Awesome ! When will the LLM model who sees you speak to you before we speak to him?

So, please give me a reason to pay for chatgpt pro...

Natalia working hard in her OF

I just passed some integral questions to solve....got wrong answers though!😐

thank you for the video:)

Nice! Could you please do a similar video about the new yi-1.5 models?

Love your videos so much! Can you make a tutorial video on FlexClip’s AI tools? Really looking forward to that!

What chat api app are you using in your demo?

Pleasee show how to create ahents on it

cant wait to see you test this deeper :)

I like your speed, no time wasted!

This video has some weird cuts. Either way, nice video.

i wonder if you would have dropped to 0 temperature what would have happened

Why not just call it GPT Omni... Everyone will see the 4o and think 40.

Doesn't it make sense to change the numbers in these questions? Didn't it train on them?

“I’m gonna add one more question” - I still hear it in my head 😂

Isn't it better to have Temperature around 0.1 for asking logic / math questions?

Let's do this

This is amazing!

First,

But i like coding! 😂 would love to see a video on directus though. That way you give people the choice to self host or cloud.

no need to worry about learning python anymore! what a pity !😃

I only see examples with PandasAI with a single table, which is unrealistic in most cases. I would like to see an example with a slightly more complex database. I tested it with the SmartDatalake instead of the SmartDataframe. Also the agent. But it was unsuccessful in most cases. One table for employees, another for their holidays. Just working with 2 tables rendered PandasAI useless in most cases. Adding (insert) data into a table was always unsuccessful, even though it said it was doing the input.

Thanks so much for the ever amazing content

Wow 👌