Really enjoyed this video. Your insights on Perplexity AI are useful. I’ve been using Myko Assistant lately, and it’s a game changer for getting quick, accurate info without the setup hassle. It feels more streamlined compared to others.
@alvintohw4 күн бұрын
How do you iterate on the Business Model Canvas? Is there a way to see the prompt or code behind. It's interesting it takes the input, goes through the LLM to generate the ideas and outputs to the canvas.
@aiunleashed5093 күн бұрын
you can make changes by just entering new prompts in the title bar. Also you can download the html file and edit from there
@theonerealestatesecret7 күн бұрын
I Created something on websim I'll try these suggestions too. Subscribed. Thank You ❤❤
@pacific78528 күн бұрын
Hi , is it imp to download the model and do the setup or just calling api in backend is enough to get ai response does meta provide the api service since i dont know python I rely on api only 💀
@aiunleashed5097 күн бұрын
No you don't need to download the model. In this video it's run in the cloud with groq(not meta directly).
@MrRossss19 күн бұрын
Also if you click a link and it doesnt create a new page you can just change the url and it will understand what you want and auto generate a new page based on the url eg it will create a new page for the link!
@aiunleashed5097 күн бұрын
Great tip!
@HomeGrowEngineering12 күн бұрын
Somehow this doesn't work anymore. Botpress keeps using its own model. Its not querying my chatgpt assistant
@aiunleashed50910 күн бұрын
Oh no, let me check it out again, haven't used bot press I a while
@noobieviolet14 күн бұрын
BOOSTING THIS BECAUSE YOU NEED TO BE BIGGER
@alissonpina533915 күн бұрын
Good job! 👏I have a question: if I want to send, for example, 4 images, what would be the correct way to do it?
@aiunleashed50910 күн бұрын
you can pass multiple images at once, see example: community.openai.com/t/gpt4-v-the-order-of-multiple-image-inputs/519966 the problem is you can't tell which response is for which image (they are not indexed the same way they are submitted). Not sure there is a solution to this yet (let me know if you find one). The workaround is just looping through the images and getting responses one at a time.
@ashleylauren413316 күн бұрын
Thanks! Super helpful and straight to the point
@mohamedalichakroun696716 күн бұрын
Thanks , but It can extract data from images ? I tried llava but it gave me a bad result Always say that the image provided doses not have good resolution ? Can you please make a video about that like you did with gpt4 for conversion img to json
@aiunleashed5093 күн бұрын
I don't think you can use this method for AI vision. Working on a video about AI vision with Claude
@Tprakh-iw6qt18 күн бұрын
Thanks a lot. It was truly helpful
@mohamedalichakroun696721 күн бұрын
Thanks MEN !!! it's so helpful, I planified to work for my own project with GPT 4 o for extraction and I'm wondering about the billing when use multiple image in single API For json extraction?
@aiunleashed50920 күн бұрын
Hey! Thanks. the way it works is you buy prepaid credits from OpenAI. Then you are charged by the tokens required to read the image and return the JSON response. The biggest variable is the resolution of the image. On this page OpenAI provides a calculator: openai.com/api/pricing/
@lifewithG-bengs21 күн бұрын
Wait what so this is completely free?
@aiunleashed50920 күн бұрын
Yes! at least for now :-)
@theknowledgejar22 күн бұрын
Maybe the site has changed but there is no download template.
@PhilShary13 күн бұрын
Yes, they removed it, which is a shame, things are now super difficult for non-coders
@theknowledgejar13 күн бұрын
@@PhilShary yup, found a solution and posted it on my AI KZbin channel kzbin.info/www/bejne/rYSzgWOVetqIjacsi=eB4SUeV6aGAt9-7i
@aiunleashed50910 күн бұрын
Thanks! will pin this so people can see the new way
@ArabmilitaryNews22 күн бұрын
very helpful, thanks!
@roofoofighter26 күн бұрын
GPT4ALL, Pinokio
@rodvik27 күн бұрын
Great vid! I would add huggingchat the huggingface customizable chatbot. The reason being you can use custom system prompts like GPT's called assistants in hugging chat. There are aton of ones to share or use or you can write your own. They are particularly useful for uncensoring models, if your business is in the arts or entertainment then being able to write and have a tool to help with grownup subjects is invaluable.
@aiunleashed50926 күн бұрын
I haven't spend as much time using Huggingface as I could like. Definitely will be exploring it. Great suggestion!
@playgirl4429 күн бұрын
when i do the 'python project.py' step in terminal it doesnt work. how to fix this?
@aiunleashed50928 күн бұрын
what error do you get?
@roberthaisfield3349Ай бұрын
Thanks for the video! I love the business model canvas generator, and an advertising quiz was a great idea.
@aiunleashed509Ай бұрын
Thanks glad you enjoyed, I really liked that quiz too. I have made a bunch more for sports and other things I am into it, they are really fun
@KoratNewsTHАй бұрын
so amazing! Thank you so much
@mauricemaeterlinck8799Ай бұрын
Amazing. Thank you. got my sub :)
@RetroTunesShortsАй бұрын
I downloaded the HTML of the Business Model Canvas Creator from Wibsim for testing. After placing the code on my Wix website page using HTML, it showed that the server IP address for the Business Model Canvas Creato site could not be found. In general, the Canvas Creator doesn't work. What else can I do to host it just for testing purposes?
@joelkaben11 күн бұрын
Upload the HTML file you downloaded to a normal web host directly. Don't use the raw HTML in site builders. To edit the HTML file, you can use Notepad++ or other similar coding editors.
@musilabАй бұрын
That's really interesting! Do you have the link for the Business Model Canvas Creator? :)
@aiunleashed509Ай бұрын
Thanks! check it out here: websim.ai/c/i7JHS1lhxhlhVhe7A
@emerson-sheaapril8555Ай бұрын
Great video, great presentation
@aiunleashed509Ай бұрын
Thanks so much!
@dlombardАй бұрын
Thanks for the insight. Also, you're just a likable guy putting good things out into the world. I appreciate that!
@aiunleashed509Ай бұрын
Thanks really appreciate it. Got some negative feedback on the video, so this really gives me some confidence to keep improving
@RyanDoserAIАй бұрын
Claude 3.5 Sonnet > GPT-4o Claude Projects > Custom GPTs Anthropic > OpenAI
@aiunleashed509Ай бұрын
I would agree! but things change fast in AI
@RyanDoserAIАй бұрын
@@aiunleashed509 Yes they do lol
@Lauren-E66Ай бұрын
Thanks for good video. Question: I try to add Midjourney Bot but Continue button grayed out. Instead, I see a note that that reads: this requires you to have manage server permissions in the server. I have gone into rules and created full permissions for myself as the administrator. Also, I have subscribed, etc., but I still can’t seem to add the bot. Any suggestions would be greatly appreciated.
@aiunleashed509Ай бұрын
Double-check server ownership: Ensure you are indeed the server owner. Only the server owner can add bots with certain permissions. You can verify this in Server Settings > Members. Review your role permissions: Even as the server owner, sometimes role permissions can override your abilities. Go to Server Settings > Roles and check the following: Your highest role should have the "Administrator" permission enabled Alternatively, make sure "Manage Server" permission is explicitly enabled Server settings: Check Server Settings > Integrations. Ensure that "Allow direct add of bots with required permissions" is enabled.
@Lauren-E66Ай бұрын
I tried but to no avail. Any chance I can hire you for a tutorial? I am utterly confused.
@aiunleashed509Ай бұрын
Hey I can take a look but won't have time till the weekend. Email me: [email protected]
@user-xd3uz4hs1rАй бұрын
i might switch to claude but i want to know if there is a limit on how much i can interact with it, i dont want to be timed out just to go back to gpt4o because i reached my limit for how many messages i can sent on claude.
@aiunleashed509Ай бұрын
Hey there is a limit. not sure exactly but for me it was usually after about 10 prompts it said wait 5 hours. The paid version has 5x the usage limit
@michaelbailey1420Ай бұрын
Claude really impressed me when it spit out an .stl file of a cube. (.stl files are for 3D printing slicers) Claude actually wrote the code of the .stl I copied it and dropped it into prusa slicer and it was perfect. it was only a cube but still very impressive
@aiunleashed509Ай бұрын
That's so cool. I want to get a 3D printer now
@AkidilCocuguАй бұрын
I use both and Claude is like Ferrari while 4o is slower and more stupid than a turtle.
@micbab-vg2muАй бұрын
the same I cancelled GPT4o and now I have Claude 3.5 and Perplexity:) only. I use OpenAI - API only:)
@uplifthabesha754Ай бұрын
'message': 'The model `gpt-4-vision-preview` has been deprecated
@aiunleashed509Ай бұрын
you can just use gpt-4o now, its multimodal including vision. Check my channel I did a video on it.
@emilsmezinskis7905Ай бұрын
why I have this error? export GROQ_API_KEY="my_api_key" export: The term 'export' is not recognized as a name of a cmdlet, function, script file, or executable program. Check the spelling of the name, or if a path was included, verify that the path is correct and try again.
@aiunleashed509Ай бұрын
its likely you are trying to run export in a windows environment. Try set GROQ_API_KEY=my_api_key...google setting windows environment variables if that doesn't work
@remisanlis5344Ай бұрын
Hi, very interesting video ! Watching from Paris, France ! I'm currently developing a solution based on that, is it possible to talk together briefly on messenger or something else ?
only big problems they have : no support - i can't upgrade card declined - many facing same problem
@aiunleashed509Ай бұрын
Thanks for the heads up. I didn't have any issues, but it's not good to lack customer support. I suspect they are getting a huge inflow of subscriptions right now, so likely understaffed for it
@spiker.c6058Ай бұрын
From my understanding, they kind of require you to enter a card issued from where you live physically, you can't really bypass that with a vpn, those who've successfully done that have seen their accounts banned shortly after.
@micbab-vg2muАй бұрын
Agree - Claude 3.5 is my new default LMM:) and it is just Sonnet.
@yulukagigantemuriel4181Ай бұрын
You're awesome!
@aiunleashed509Ай бұрын
Thanks, comments like this keep me going. You're awesome too!
@SunilSamson-w2l2 ай бұрын
Does it also work with PDF files instead of .img ?
@aiunleashed509Ай бұрын
it says: We currently support PNG (.png), JPEG (.jpeg and .jpg), WEBP (.webp), and non-animated GIF (.gif). So no you would have to convert it first
@SunilSamson-w2lАй бұрын
@@aiunleashed509 thank you !
@heshamsoldaten2 ай бұрын
how to get API key ?
@aiunleashed509Ай бұрын
Look in the account setup section from OpenAI: platform.openai.com/docs/quickstart
@devamsanghavi17872 ай бұрын
Thank you so much! I still encounter some issues like I'm uploading an Invoice and every time it gives different vendor name(upper case, lower case) and how to mention date format in JSON Schema? it always return different format. how can I prompt this?
@aiunleashed509Ай бұрын
For the date format, you can prompt for it, I have had good luck. For the json schema, when the "format" is set to "date" to specify that the string should be in the date format according to ISO 8601 (YYYY-MM-DD). I think you can also specify pattern for different format. I haven't seen the lower case upper case issue. Only idea would be make it so each word had capital letter in python in post processing (for consistency), although names that are all uppercase would be a problem...
@adrielomalley2 ай бұрын
Appreciate a new face ;-) let's see this channel grow!
@theotherguy61552 ай бұрын
I can see how this looks great, but if you understand LLM under the hood it's not giving you "recommendations" as much as it's approximating what an acceptable response would be. This is ok if you "don't know where to start and want to do something" but as far as actual suggestions i don't think it's any better than you taking 5 minutes and sorting those columns in excel. lol
@aiunleashed5092 ай бұрын
Fair points. The first example was straightforward, but the idea is that you can build on this for more in-depth analysis. I believe the insights gained this way can be far superior to what an average business owner could achieve by merely sorting columns etc. I’ll aim to use more complex use cases in the future.
@E79ric2 ай бұрын
Hi, Thanks for this video on using the GPT-4o Vision API. I'm using the code shown to detect text in images, and it's working very well. However, when I request the pixel coordinates for sections of the invoice (general information, product details, and payments), the accuracy is not very good. Could you provide some advice or demonstrate how to improve the accuracy of the pixel coordinates for each section in the image? I need to locate specific areas like the invoice number, client information, tax ID (CIF or NIF), product details, and payment information such as the total amount and VAT. Thanks in advance for any help!
@aiunleashed5092 ай бұрын
Hey! I think current general purpose AI models with vision will struggle with this kind of zonal OCR. Have you tried Google Vision? it has more established history and might work with giving coordinates. I will check it out for you though, see what my results are.
@E79ric2 ай бұрын
@@aiunleashed509 Perfect!
@E79ric2 ай бұрын
@@aiunleashed509 Thanks for the suggestion. I really appreciate your willingness to check it out for me and look forward to seeing the results you get.
@rajmandaviya8582 ай бұрын
Hey, I was doing the same thing before i found your video but adding response_format was a great help. Thanks! Now i am finding a way where i send multiple images to gpt4o and get an indicator if image is rotated(it does not work on rotated images) now when it comes to multiple images i need an identifier of them to rotate required image only, Do you have anything in mind?
@aiunleashed5092 ай бұрын
Hey! Glad you found it helpful. For rotating an image, I would actually just put in a pre-processor step to do this without gpt4o. Try the auto-rotation and deskew functionality in something like tesseract. I just say that because its already a solved problem, and you could do it for free without AI, and I'm not sure the models would be that reliable
@PalkiaOwnsYou2 ай бұрын
Hi, Is there a way for the code from this video to be posted?
@aiunleashed5092 ай бұрын
Yes, here it is: github.com/AI-Unleashed/GroqResume
@HMaxTube112 ай бұрын
Good summary tour.👏🌟👍Subbed.
@aiunleashed5092 ай бұрын
Thanks for taking the time to comment, appreciate you
@BlackStudies2 ай бұрын
Anything yet on how these perplexity pages will rank? Will the search engines even crawl it?
@aiunleashed5092 ай бұрын
I doubt google would include these, but not sure
@danielalbano2 ай бұрын
Hi. Do you know the limit of tokens i can use? Im trying to transcribe an image with a lot of text, but it it stops in the middle. it seems the maximum of tokens i can use is around 1000.. How can i set more tokens per request?
@aiunleashed5092 ай бұрын
at 2:30 in the video where I set max_tokens, the limit for GPT-4o is 128,000. try a few thousand first to see if it transcribes more of the text
@thewebstylist2 ай бұрын
Amazing, just wish we could embed in our own websites
@SnakeCaseGuy2 ай бұрын
I just have a basic request for Perplexity. There is differnce between putting the search bar at top vs at the bottom is great. Its the difference between a search engine and chatbot. I don't want a chatbot.
@aiunleashed5092 ай бұрын
in the case of Perplexity it really is both an AI chatbot and a search engine. might be a nice user setting if you could choose where you want the input bar...