Тёмный
Mervin Praison
Mervin Praison
Mervin Praison
Подписаться
Mervin Praison
OpenAI DevDay Insider Secrets Revealed
12:52
2 часа назад
Llama 3.2: Best Multimodal Model Yet?
5:27
4 часа назад
OpenAI o1 in ChatLLM & more
3:12
14 дней назад
Solar Pro: Did it Pass the AI Tests?
7:41
21 день назад
How to Make Lightning-Fast AI Apps
7:33
21 день назад
Cut AI Costs with Smart LLM Routing
7:16
Месяц назад
Комментарии
@Temujsr
@Temujsr Час назад
Hello, it's impressive but can you make a guide with local images not image url?
@ramanefira
@ramanefira Час назад
thanks for the tutorial
@shahariarhossainremon6500
@shahariarhossainremon6500 2 часа назад
Hi! Can you suggest me multimodal and also "multi-lingual" LLm names, so that I can fine tune in my own language data.
@devrajgothi5149
@devrajgothi5149 5 часов назад
I am really new to this. My question is the embedding is really slow, is there some way i can use my dedicated GPU for embedding?
@nadantilttlia
@nadantilttlia 7 часов назад
not classifying medical dataset accurately.
@MervinPraison
@MervinPraison 6 часов назад
Did you try training?
@nadantilttlia
@nadantilttlia 5 часов назад
​@@MervinPraison Yes, sir. I have tried using the IU-Xray and MIMIC-CXR datasets and am still working on how it performs.
@MervinPraison
@MervinPraison 5 часов назад
@@nadantilttlia Hmm Keep the epoch and batch size as auto
@rodelsecreto9440
@rodelsecreto9440 10 часов назад
Thanks for the tutorial.
@60pluscrazy
@60pluscrazy 10 часов назад
Great 🎉
@jim02377
@jim02377 14 часов назад
I tried to make it work but I am using Visual Studio code on a windows 11 box. I ran pip install inside a python virtual environment and then ran the code. I get a path error that I don't understand. I assume this was done on a linux box and not in a python virtual environment.
@MervinPraison
@MervinPraison 5 часов назад
Probably you are using windows , try using set instead of export
@instinct1098
@instinct1098 21 час назад
Will the Llama 3.1 8B model work on Macbook AIr M1 2020?
@MecchaKakkoi
@MecchaKakkoi 22 часа назад
Nice tutorial! But what should go in the `embeddings.yml` file mentioned in the last section?
@____________________________.x
@____________________________.x 23 часа назад
2/3rds of this was generating poems. I was just here for the image analysis, so this was a bit unclear.
@nico_betancourt
@nico_betancourt День назад
What is the application that you are using to test and develop the fine tuning? Thanks! What amazing info!
@KuldeepDaftary
@KuldeepDaftary 19 часов назад
I think OpenAI's fine-tuning feature you can find is on OpenAI's API section
@nico_betancourt
@nico_betancourt 17 часов назад
@@KuldeepDaftary Yes thanks, but I want to use some models in thtat way with that interface, not just for fine tuning.
@Idiot123009
@Idiot123009 День назад
Can you create video for calling external API to call with llama3 2 Last time you created that but. Because llama isn't good to call function/tools intelligently. Because when we pass tools it always call tool. What if the query isn't about to call tool. Query is about to call the tool with none. But it always call the tool. Even passing tools_choice="auto" It's not intelligent here. Help me to sort that problem.
@nps798
@nps798 День назад
so do i need to pay for searching ? it seems like it can serve as free alternative to google search api for my RAG app right ?
@xhridhar
@xhridhar День назад
Real-time API is super expensive 😔
@MervinPraison
@MervinPraison День назад
True
@yotubecreators47
@yotubecreators47 День назад
Wow this is super informative video I wish if all RU-vid videos would be like this Thank you so much
@claudioagmfilho
@claudioagmfilho День назад
🇧🇷🇧🇷🇧🇷🇧🇷👏🏻, A good question to ask would be when we can expect real-time video, since this was supposed to be part of ChatGPT Omni release, but it hasn’t been rolled out yet.
@Yaw_Dev
@Yaw_Dev День назад
"This is amazing!" Mervin's favorite intro.. lol Greate video..
@MrWizardGG
@MrWizardGG День назад
I AM THE SIGMA and Ai is so scary
@thetrueanimefreak6679
@thetrueanimefreak6679 День назад
Im the sigma that's says tell me more >:)
@60pluscrazy
@60pluscrazy День назад
Great 🎉
@Uconnspartan
@Uconnspartan День назад
Glad you reported on it.
@tecnom7133
@tecnom7133 День назад
le Why you test it with QR code , it's supposed to read it ?
@brianWreaves
@brianWreaves День назад
Cheers for offering concise summaries... really is impressive how much you can convey so quickly. 🏆
@jlyunior
@jlyunior День назад
This is amazing ! 🎉
@stas4000
@stas4000 День назад
Wow amazing! 🎊 📈
@younginnovatorscenterofint8986
you made the paper easy to understand .thank you
@Lamoboos223
@Lamoboos223 День назад
How to remove the restrictions? Will fine tuning the model help?
@epokaixyz
@epokaixyz День назад
You've found the right comment! 🎉 1. Explore the capabilities of Llama 3.2 Vision by trying it out on your own computer. 2. Use Llama 3.2 Vision to get design suggestions for your living spaces or declutter your home. 3. Remember that AI models like Llama 3.2 are still learning and have limitations, especially with complex CAPTCHAs or finding specific individuals in images. 4. Utilize Llama 3.2 to extract data from tables within images for digitization purposes. 5. Experiment with Llama 3.2's ability to generate basic HTML and CSS code from design mockups as a starting point for web development. 6. Stay updated on the advancements in AI vision and explore the resources available to learn and experiment in this field.
@nielsstighansen1185
@nielsstighansen1185 2 дня назад
Watched the video and totally missed the shocking part 😂
@tecnom7133
@tecnom7133 День назад
hhh
@agentDueDiligence
@agentDueDiligence 2 дня назад
Hi Mervin, What do you think about costs. Let’s say you have a simple vision model and it works for your usecase. Is it cost wise effective to host this LLM yourself and switch out the usage of OpenAIs models? Do you have any indications which you can share?
@Idiot123009
@Idiot123009 2 дня назад
What if not all the query needs to call the tool?
@brianWreaves
@brianWreaves 2 дня назад
Great explanation of the distinction with/without LMStudio! 🏆
@DaveVT5
@DaveVT5 2 дня назад
What is the UI at 2:01? Looks like you may be playing a video provided by Meta but I'd like to test the multimodal capabilities of the model and I don't think Open WebUI supports it.
@hbdaisy6955
@hbdaisy6955 2 дня назад
Could I choose different voices?
@benedickphilmonor5859
@benedickphilmonor5859 3 дня назад
please help i cant seem to open index.html
@HelloFriendz657
@HelloFriendz657 3 дня назад
Please there is need for videos on ,how to deploy these ai agents created in vs code or pycharm to production and websites, eg crewai .I do know whether people feel same way, how will your client receive your code in maintainable stable format.
@ghalianas9211
@ghalianas9211 3 дня назад
Is this free ?
@AjayGupta-dx7wg
@AjayGupta-dx7wg 3 дня назад
@Mervin: Any specific reason to have chroma db instead of more powerful vector db available?
@aimademerich
@aimademerich 3 дня назад
Phenomenal, anymore LLMs like this?
@aimademerich
@aimademerich 3 дня назад
Phenomenal
@MariaThompson-d7y
@MariaThompson-d7y 4 дня назад
Hall Robert Lopez John Allen Daniel
@KustieY
@KustieY 4 дня назад
no 11b so far?