วีดีโอ

vectordb's i think Scaling Gets Pricey Sure, they’re awesome for AI and similarity searches, but man, handling huge datasets means hefty costs-both in hardware and operational overhead. Complex Setup They’re not like your typical database. Setting up and managing a VectorDB requires some deep AI and indexing knowledge. Not exactly plug-and-play. Niche Use Cases Great for finding similar images or text, but for regular old-fashioned relational data or transactional queries? They’re pretty useless. Latency Issues When you’re dealing with really high-dimensional data, search times can drag unless your hardware’s top-notch. Interoperability Headaches They don’t integrate well with traditional tools, so getting them to talk to your existing systems can feel like pulling teeth. Garbage In, Garbage Out If your embeddings (the data fed into the DB) are off, then your searches are gonna be garbage. There’s no room for error there. Still Evolving They’re trendy, but the tech is young, and standards aren’t fully baked yet. So, investing now could be a gamble if a new shiny tool comes out next year. Basically, they’re like that fancy sports car-awesome for the right track, but terrible for hauling groceries or a road trip with potholes.

Huggy wuggy huggy😂😂🎉🎉.

Great content, as always! Could you help me with something unrelated: I have a SafePal wallet with USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). What's the best way to send them to Binance?

Really enjoyed this video! This really gives a clear view of what’s coming in the crypto space. Curious to dive deeper? Feel free to check out my bio! Thanks for the video, I'll be watching for more updates!

When I try to follow you up with Grounding, I face this error all the time Failed to submit prompt Error message: "Request contains an invalid argument." Status: 400 Error code: 3 Request ID: 4734487872953052040 I went through all the possible walkarounds for this on various forums but there is no working answer on this. Your help would be greatly appreciated Thanks.

this is amazing, thank you so much

indians 🇮🇳🇮🇳🇮🇳🇮🇳

Hi, can you share the github or colab to replicate the examples?

Hi, I am trying to use context caching in cloud shell for our project. But I cannot import caching "from vertexai.preview import caching" getting import error. How to get rid of this error? In some other videos in labs they are importing using from google.generativeai import caching. Which method in need to use and why am I getting this error. 37:15

frequentist vs. bayesian view=fixed mindset vs. growth mindset

Why does the Bayesian AR match the training data so much better than the frquentist ARIMA?

This is amazing!

Nice video

Your video is the first one I've found that answered the simple question of what a repository is and how many projects to add to it. I know that sounds foolish with all the info on the web, but my head is so full of new info that I haven't even known what search terms to use. I'm putting all my initial, simple projects in one repo to get my portfolio started and then moving on to complex prjects with data that needs cleaning and web scraping. Thank you so much, Saishruthi!

Thank you! Great presentation

It is really clear and good video. I really recommend this video!!

Nice speech

Hi! Will you be able to upload the Austin meetup from 4/24/2024 with enterprise architecture?

Masterful communication and presentation skills, damn!

Thank you!

Great video!

Can you please share the link for Google drive code ?

4 years late here

extremely helpful, thank you sir!

Just leave!

🌹

is it possible to follow this along with the azure free trial version? thanks

Thanks for recording and uploading these to TH-cam. Much appreciated !

Thanks for recording and uploading these to TH-cam. Much appreciated !

This is by far the most intuitive explaination of transformers i have come across. Will be referring people to this , thanks.

This is really clearly explained

I wish they taught me this way at university...

How to enroll for this course, any prerecorded videos available?

ApertureData slides from the event: drive.google.com/file/d/18elQAzrhD61F-n7vWZUSyazsOmEVDEXz/view?usp=sharing , showing examples of what's needed from data architecture when setting up production scale ML pipelines, including scalable vector database, efficient graph-based filtering, and high performance multimodal data management.

How can you load the data file from Kaggle?

Thank you quite insightful :) 😁

this is not design patterns! design patterns refers to patterns like strategy, builder, observer, factory, etc. u should have talked about how those patterns can be applied to make code more extensible flexible and debuggable.

🤷 *Promo SM*

I signed up for this talk but didn't go. I am glad that I actually watched the recording (something that I normally won't do, who's got the time when there are tons of videos to watch?). I learned a lot from it. Thx.

Aric LaBarr is great

Thank you. Very much appreciated

I was there. Excellent presentation. Second half on Multimodal Embeddings starts at 33:10. There is a quiet part of the video, you need to hop to 37:10. The video effectively ends at 49:49, it goes quiet after that.

Shouldn't we add bos and eos tokens on the dataset samples?

Hello, it's possible to use Autopilot to classify images in jpg? Or it's necessary to convert our images to an array like Numpy array?

My question was "What's the difference between Beam and Flink? And why do we need Beam if there is a Flink?". This video helped a lot. Thanks !!!

Good vedio ❤❤

We have an issue where beam combined with tensorflow transform (i.e. AnalyzeDataset) tends to crash by out of memory errors when working with a lot of columns, say more than about 520. We split them, tried preprocessing only parts and all parts worked on their own, but as soon as we put in all columns, it becomes incredibly slow and crashes. Any experience with that, any help?

excellent presentation @Aric LaBarr.... well structured and super clear!

Oh man, I missed this by a week...Definitely would have made the trip across 520 if I had stumbled upon this event info in time! I can't believe the topic is exactly the app I've been working on at the moment too which seems like a wild coincidence only adding extra insult to my already bad timing! 🤨

Hey guys, awesome work Sorry to say, but I think it would be best to make sure the speakers have a clear mic This video is pretty clear, but the recent ones are pretty difficult to listen to Having clean mic input is pretty easy to set up! Cheers