I'm attending React Advanced Conference 2024 - get your free remote ticket and join me there with 5k other engineers and 50+ great speakers
https://t.co/YwrD5utvS5
#ReactAdvanced
Semantic Search Engine for Images: find images from a database using natural language descriptions.
Team: Sandhya Govindaraju, Utkarsh Vardhan, Gabriella Chiribau, Amit Kumar Sah
Really impressive monitoring setup too. Try it for yourself!
https://t.co/QSvVmWBCUE
One needs a form of feedback. In school, we had tests and contests. Today, there are edX classes. The probability class kept my spirits high.
"If you didn't do calculus in 2 years or so..."
30 years, yet I remembered.
"It's been so long,
But it's all coming back to me now."
:)
How do you... study math and like... learn anything? Asking for a friend.
In all seriousness, how do you keep your spirit up when you self study amd keep hitting a wall every other step?
That's how I learned. Built an ubuntu machine with a 3090, docker and different stack components to understand end-to-end.
Even with a good math background, stats and "theory" were a pretty heavy lift.
Probably a popular opinion: The first course in an ML education should teach a basic end-to-end workflow (raw data to automation). It��s so much easier to learn math, stats and theory when they are pegged to the components of the overall system.
After coaching a lot of very successful people, here's an unsexy truth:
A lot of success is learning to grind when it's boring.
Staying engaged, working, and seeing it through when it's "boring" is a skill. If you expect all building to be exciting, you'll never make it.
Among learners there are four defects with which the teacher must make himself acquainted. Some err in the multitude of their studies; some, in their fewness; some, in the feeling of ease with which they proceed; and some, in the readiness with which they stop. (BR, Xue Ji 10)
Math competitions like AMC, AIME, and the USAMO were formative parts of my middle- and high-school experience.
I never quite made it to the IMO, but our new neural network is gearing up to be a formidable competitor at each level of the math competition circuit:
It is a truth universally acknowledged, that someone in possession of an rtx3090, must be in want of #TensorFlow2.
I added tf2 support to #AmpliGraph and its hands-on intro to #KGE: link prediction, clustering, and #tensorboard projector data.
https://t.co/PHGWtbiawU
#inprogress
It's amazing that a simple idea, such as the slope of a tangent line could have so many incredible consequences.
Modern calculus was developed ~360 year ago.