Video yükleniyor...
Video Yüklenemedi
🚀 Just built this cool carousel! - Background gradient by DevL∞p's Nitro Palette - Gradient blur inspired by Brice Zele's React Native Gradient Blur And yes, Expo is the backbone making it all possible Had to try it out. Here it is! 🙌
13,326 görüntüleme • 1 yıl önce •via X (Twitter)
11 Yorum
@BriceZele @expo Be sure to check out Nitro Palette 👇
Absolutely in love with the review from @xzeldawaifux 😍 She’s rocking it in our ambassador program and earning up to 20% commission on every sale! 💰🎉 Wanna join the fun? Sign up for FREE at 🌟 Drop your unique promo code in the comments 📢💫
@BriceZele @expo And this cool repo as well 😎
@endlessFrontend @BriceZele It’s beautiful 🥹
@endlessFrontend @BriceZele Thanks a lot! 😊
@endlessFrontend @BriceZele @expo I usually use expo linear gradient package with react-native-easing-gradient package to blend images towards the bottom. Going to try the gradient blur method you posted and see how that compares!
@endlessFrontend @BriceZele @expo Sure! Do checkout, I just found out about it while I was working and it turns out awesome! 🚀
@endlessFrontend @BriceZele @expo Very cool! Looks amazing!
@endlessFrontend @BriceZele @expo Glad you think so! ❤️
@endlessFrontend @BriceZele @expo Does it work on Android?
@endlessFrontend @BriceZele @expo Yes Sajjad!
![CSS Tip! 🤙 You can use mask-composite and some JavaScript to create this pointer proximity following glow border ✨ .glow { mask-composite: intersect; mask-clip: padding-box, border-box; mask: linear-gradient(#0000, #0000), conic-gradient(#0000 0deg, #fff, #0000 45deg); } The trick is to mask a background-image with a combination of mask layers. mask-composite: intersect; means the mask used will be the intersection of the layers 🔥 use source-in, xor; in browsers that don't support intersect; In this demo, you can use pseudoelements and rely on scoped custom properties to do a lot of the heavy lifting for you 🙌 Once you've masked the background, you need to update the starting angle of the conic-gradient on pointermove 👆 You can work that out by getting the center point of each card and then calculating the angle between that and the pointer with Math.atan2 🤓 let ANGLE = Math.atan2( event?.y - CARD_CENTER[1], event?.x - CARD_CENTER[0] ) * 180 / Math.PI ANGLE = ANGLE < 0 ? ANGLE + 360 : ANGLE; CARD.style.setProperty('--start', ANGLE + 90) You plug that into your conic-gradient mask as a custom property accounting for --spread ⚡️ conic-gradient(from calc((var(--angle) - (var(--spread) * 0.5)) * 1deg), #000 0deg, #fff, #0000 calc(var(--spread) * 1deg)); To get the blur, you apply a blur to the glow container on each card 🤙 .glows { filter: blur(calc(var(--blur) * 1px); } That's it! Layers of masks that are clipped and composited before being blurred 😎 The added trick is to fade each one in when the pointer is in the defined proximity of the card. For example, don't show unless within 100px of a card. You can see that in the video. Check out the JavaScript code for that 🫶 Couldn't resist making this one 😁 CodePen.IO link below! 👇](https://image.24vids.com/tw-1734369933558010226/ext_tw_video_thumb/1734365662590418944/pu/img/320Dgq_k3RZoZOFV.jpg)
![Backpropagation by hand ✍️ ~ 11 steps walkthrough below Backpropagation is the algorithm that actually trains a neural network, and it is where most people stop following along. It is not calculus you cannot do. It is matrix multiplication, working backward, one layer at a time. So I drew and calculated one entirely by hand. Goal: push the loss gradient back through a 3-layer network and land on a new value for every weight and bias. = 1. Given = A 3-layer perceptron, an input X, predictions Ypred = [0.5, 0.5, 0], and the truth Ytarget = [0, 1, 0]. = 2. Backprop gradient cells = Let us draw empty cells for every gradient we are about to compute. The shape of the answer comes first. = 3. Layer 3 softmax = We get dL/dz3 straight from Ypred minus Ytarget = [0.5, -0.5, 0]. No chain rule needed, and that shortcut is the whole reason softmax and cross-entropy are paired. = 4. Layer 3 weights and biases = Let us multiply dL/dz3 by [a2 | 1]. One multiplication gives the gradient for W3 and b3 together. = 5. Layer 2 activations = We multiply dL/dz3 by W3 to get dL/da2. The gradient moves back across a layer the same way the signal moved forward. = 6. Layer 2 ReLU = Let us pass it through the gate: keep the gradient where the activation was positive, zero it everywhere else. = 7. Layer 2 weights and biases = We multiply dL/dz2 by [a1 | 1]. The same figure as step 4, one layer up. = 8. Layer 1 activations = Let us multiply dL/dz2 by W2. = 9. Layer 1 ReLU = We apply the same gate again, now on a1. = 10. Layer 1 weights and biases = Let us multiply dL/dz1 by [x | 1], and every weight in the network now has a gradient. = 11. Update = We subtract, and the network has learned. In practice a learning rate scales this step. The gradients: dL/dz3 = [0.5, -0.5, 0] dL/da1 = [1, -2, 2, -1] dL/dz1 = [0, -2, 2, -1] The takeaway: matrix multiplication is all you need. Just like the forward pass, backpropagation is matrix multiplications end to end. You can do every one by hand, slowly and imperfectly, which is exactly why a GPU's ability to do them fast mattered so much to deep learning. 💾 Save this post!](https://image.24vids.com/tw-2079208495094677511/ext_tw_video_thumb/2079208483669315584/pu/img/dDD0x6lHYHuV1POQ.jpg)
