so this app was from the show Silicon Valley? I was asked to design a hot dog classifier app in a startup interview and didn’t catch the reference at all
Did we actually use the model to predict the images at the end of the video? Correct me if I'm wrong, but it looks to me that we just showed the images and labels from the validation dataset, instead of using the model itself to predict the labels
So if computers can differentiate objects in an image, why is that still used as security measures to determine whether an individual is a human or a bot?
Hmmm - Lots of good stuff here, but about minute 48, the screen goes black. My guess is whatever was being displayed was blanked out due to the application's author complaining to TH-cam. My suggestion is to redo the segment and repost it if it's an important enough.
Hey Kylie, big fan. Quick question, why aren't we using an ImageDataGenerator for data augmentation? With this method you have a variety of techniques and it takes less amount of code. Loved the video. Thanks a lot
sure import numpy as np import pandas as pd import tensorflow as tf import random import matplotlib.pyplot as plt from tensorflow.keras import datasets,layers,models import tensorflow_datasets as tfds ds,ds_info = tfds.load('food101',shuffle_files=True,with_info=True,as_supervised=True)
Jin Yang's legacy lives on... can't wait to deploy my Not-Hotdog app
Top comment
so this app was from the show Silicon Valley? I was asked to design a hot dog classifier app in a startup interview and didn’t catch the reference at all
@@philipbutleryup 😂
😂😂
Octopus
For those who wonder how to get the missing code at 49:19 just click on the colab link in the description
JIAN YANGG!! ~ eric bakman
There's a section of the video where her screen is not shown. Just blank! 49:52
Same here!
lmk if u figured out the code! Thanks
We're adding a note about this. The code comes back at 49:18
Jin Yang!!!!!
Bruhh..silicon valley
Kylie Ying is one of my favorite lecturers. Her speech and explanation are nice to hear.
you seriously should make a full course about computer vision your teaching is pure gold , it is a steal for making this for free
Every single Gilfoyle in this Earth right now: lol
Kylie Ying is a genius! I'm just watching her videos to learn python. She teaches very well.
no. she the fk watches left monitor and just copy it
@@twinkle-d7b That may be, but it is done for didactic purposes. I don't think she doesn't know what she's doing.
Hotdog and pizza is code for something else, they say.
Yesss finally someone who can see through their BS. Either that or we are extreme conspiracy theorists 🤣
Now this is how everyone should teach programming! Now I want a hot dog!
"Jin Yanggggg!!!!!!"
Jin Yang, I'm not gonna yell
Finally something on CNN
In freecodecamp there are already many previous contents on cnn.
wonderful course, I hope there is a plan to make videos about robotics and machine learning and deep learning
Now make it figure out if hot dogs are sandwiches
Did we actually use the model to predict the images at the end of the video? Correct me if I'm wrong, but it looks to me that we just showed the images and labels from the validation dataset, instead of using the model itself to predict the labels
I wish she had used PyTorch.
I ran into error after 57:13 i can't continue can someone help
I am facing an issue while downloading food101 dataset. The download button of the website is broken. Is there any other alternate link?
Como pode existir um conteúdo de tanta qualidade grátis no TH-cam?
So if computers can differentiate objects in an image, why is that still used as security measures to determine whether an individual is a human or a bot?
You are actually providing data yourself since you're better at image classification
When you select the images in a CAPTCHA, you are helping train some model.
Jin Yang😆
you the best, if i have to compare this video, i will compare with Andrew Karparthy.
Thank you! Best explanation of cnn's i have heard so far.
what a perfect way of teaching thanks kylie Ying
Why I can't save this video to my playlist? 😢 Thank you for the content.
JIAN YAAAANG!
What prerequisites do i need to have to understand this video?
Hmmm - Lots of good stuff here, but about minute 48, the screen goes black. My guess is whatever was being displayed was blanked out due to the application's author complaining to TH-cam. My suggestion is to redo the segment and repost it if it's an important enough.
Screen capture is not working starting @49:25 and doesn't work for 3 minutes.
We are adding a note about this. The code comes back at 49:18
Fun Fact: she didn't release her AI course.
Erik backman smoking intensifies
would love to pay for ai / ml bootcamp by you
this is gold
Tonight gonna be a looooong night 💪🙏
With 25% drop before your output layer your theoretical maximum accuracy is 75%.
wow now i know how much I still have to learn!
Is it a black screen on purpose at @50:42
JIAN YAAAAAANG!!😂😂
Young Jonathan 😂😂😂
Hey Kylie, big fan.
Quick question, why aren't we using an ImageDataGenerator for data augmentation? With this method you have a variety of techniques and it takes less amount of code.
Loved the video. Thanks a lot
49.5 to 52.5 minutes, there is blank screen. can anyone help?
one of the best episodes of Silicon Valley
could someone please share the lines of code starting on 49:18 please?
sure
import numpy as np
import pandas as pd
import tensorflow as tf
import random
import matplotlib.pyplot as plt
from tensorflow.keras import datasets,layers,models
import tensorflow_datasets as tfds
ds,ds_info = tfds.load('food101',shuffle_files=True,with_info=True,as_supervised=True)
Just a reminder, still waiting for that Reverse Engineer and Malware Analysis course. 😅
Learning more in 1:30 hour more than in entier university course. Grat thank you
Just remember the scene from silicon valley series🤣🤣🤣
You guys are great, thanks so much for this content!
Provide source code
JIAN YAAAAAAAAAAAAAAAAAAAAAAAANNNNNNGGGGGG
I love the fact that she shared the slides.
52:42
Taking more data and increasing the layers will increase the accuracy.
Up to a point.. well, technically yes.. Over fitting.
no. it won't. in this case the dropout(0.25) will flat out give wrong result 25% of the time even on 100% accurate model.
your lesson make me hungry :))
boa steve jobs de guapituba
Thank you for all that you do
HEYY!!! … You did a nice job
No views?
Very well explained
29:52 / 1:27:41
Super!
YAAAAS
Hahaha nice 👍🏻
Thanks...
why people still using TensorFlow?
What is wrong with TensorFlow?
Pytorch ftw
😍
❤❤❤
👍💯
This is so dope!
man I'm 14 and iI don't know if I should be a game developper or a full stack developper
I’ll date you. I live in Tokyo. ❤ Friend me.
Cant wait to upload my hotdogs