TensorFlow.js Quick Start
Vložit
- čas přidán 2. 04. 2018
- Get started with TensorFlow.js by building a digit recognizer from scratch in this quick start tutorial angularfirebase.com/lessons/t...
TF.js docs - js.tensorflow.org
Machine Learning Crash Course - developers.google.com/machine...
Kaggle - www.kaggle.com/jeffd23
MNIST - ml4a.github.io/demos/confusio... - Věda a technologie
Welcoming TensorFlow to the web - tons of potential for web developers with this library. Let me know if you want to see more TensorFlow.js content in the comments.
Yes of course, I am really excited
Amazing stuff guys.
Awesome, continue!!
Yes we need with more real time example
yes... we need more tutorial. thank you
0:22 most educational part of the video, i learned how to write 4 in 1 stroke
😂
Should absolutely do more tensorflow videos, very cool stuff.
THIS VIDEO IS SO DIFFERENT AND somewhat nice than the fast paced videos you make today
I just saw the keynote and here you are with a tutorial. You are really awesome.
+Kiran Nice, looks like there were a lot of interesting talks I still need to watch.
Awesome! I've been watching videos/tuts on TensorFlow all day! So, Perfect Timing!
Sweet! The tf.Sequential API is especially cool because it's similar to Keras - my all time favorite Python package.
thank you, we need more of TensorFlow.js
This was really interesting. Would love to see more tensorflow.
"But be careful, data science competitions can be very very addicting"😂😂😂
Famous last words... ;)
😑
This channel 🔥💯
Watching after 6 years😂
I am so super excited because i have been looking forward to tensorflow.js. Thanks for your vid.
Glad to hear that. It's going to spark some really cool web apps in the near future.
That was a great intro vid!!! TFJS is very exciting as JS is the language of the web!
This video is still as good as gold even after 4 years.
More tensorflow.js 🙏
Great tutorial !!! Please add more stuff for beginning to learn how to implement Neuronal Networks solutions
I would love to see more of tensorflow.js
Linear regression is supervised learning right..? So that trained sequential model comes with tensorflow or did you import that from keras as well..? (Btw, never knew you were crushing it in kaggle before entering the firebase space.. This was a great vid.. From the way you narrated, seems like you are delighted about this tensorflow.js and the whole new world of possibilities it provides right in the browser.. :) :) )
Thanks Raja! Yes, when you have labeled data it's supervised learning. The Sequential model was not pre-trained, I trained it in the browser by calling.fit(), but doing a linear regression with TF is overkill.
haha i love how politly you pointed out "because you're MOST likely a human being.."
This just made my year. Please do more ML-focused videos. And thanks for all the Angular videos so far.
Thank you. Will do!
Awesome. Need more ML video like this with js and angular
I agree, I have some pretty interesting ideas for browser-based ML.
Thanks. Your tutorial is very good. I follow all you video and blog. I am a js developer from India. Carry on your best work. 😊👍
more with tensorflow.js please !
Thank you for sharing this informative video! 🐯🖐🏻 I hope you are well.
Thanks for this amazing video!
I need to start wayyyyy before this.. Lol.. Nice video.. I'll be reaching this level shortly..
I am so happy right now I can't even tell
Jeff is my favourite tech youtuber!
Thank You for Great video....
keep going bro your videos are awesome
Very interesting! Thank you!
helpful and well explained, thanks!
Amazing video! Everything was well explained. Thanks! 😄😄
I want to see more Tensorflow JS content
Could you make a video of how to participate in a kaggle competition, get the datasets, and submitting it? Please
I want to train the model in Python and then use it in a web app 😁
True: Tons of potential indeed!
this is the only useful tutorial on whole youtube
cool! do a series on this!
Well how much mathematics I need to get started with machine learning and deep learning!
And thanks for such a great content!
Just have a high-level understanding of liner algebra for matrix multiplication and calculus for gradient descent.
That is so cool,a front end for ml...
That's really cool sruff.I think sonner the AI code will be written in Angular
I'm creating a Students' Performance Prediction system, I want to use tensorflow js. Any advice?
More content on tensorflow
Kinda useful, I do made a auto encoder to denoie images I'm still learning these kind of things
Have you any tutorial about Voice Recognition in Tenserflow?
I love your channel ❤
+Suraj, thank you for watching
Thanks for your great videos.
Is it useable (in terms of performances) in an Ionic app?
as of now, yes - newer versions of webview(what runs your ionic apps on the mobile device) support webgl(gpu api for browsers to make your models run faster) - that said, you should use mobile optimised nets - e.g. mobile net in order to make it efficiently run in a resource-constrained environment
Wonderful 👍
We need more
Well, I'm cracking my head for too much time by now.
First: thank you for the tutorial, you were very clear, even though the documentation on this library is *so so much obscure*, even for me who have background in Machine Learning and the Linear Regression method.
I want to know more about what goes under the blanked in the fit() and predict() functions. Their arguments and return types are counterintuitive for me. And I can't find further documentation unless I learn Python :(
I did the example exactly like in the video, but the result doesn't seem to reflect a correct prediction at all. In fact, it seems random much of the time. Has someone got correct predictions with the example in this video? I did a simple graph and inserted dots manually to make up the training set roughly like a y = x function, and sometimes I get a y = -7x, or y = 5x, or some gigantic positive or negative number.
ce set me rappel mes débuts sur le jeu avec l atwork de yasuo élu et les synergies bien meilleures que les actuelles, j'espère juste que toutes les compo auront du potentielles et pas juste une à deux, ça dégouterai vite
Is it even useful if you learn Django-Ajax or Flask-Ajax calls ?
Do you know how can we translate the audio using tensorflow.js.?
Can you please show us how you connect the drawn image to the typescript code
i installed tensorflow/tfjs and it added 23 packs from 37 contributors and audited 25 packs. are these numbers correct? btw, i am learning Angular first atm .
awesome!!!
Is it a good idea to run tf.js on cloud functions ? I know predicting would be easier, but what do you think about training ? It'd be really cool to trigger training on real data whenever necessary!!
It's not supported in Node yet, but I am really looking forward to this possibility. It would be best for predicting or possibly training small models, and tightly integrated with your Firebase data.
1:45 “As you’re most likely a human being”
Oh no he got me
Ty
where can one get to know the pre trained models ?
This might be an odd question, but at times it seems as if you are able to store multiple items on your Clipboard and paste them in in a specific order. I know that you can keep multiple items on your Clipboard, but how can you paste them into the document one by one?
Get that question all the time. When recording I delete lines in reverse order, then hit ctrl-z
Angular Firebase Awesome! I was observing the added, deleted, modified line hints in VS Code and yet I didn't quite catch the fact, probably didn't pay enough attention. I'm looking forward to more videos on this topic and perhaps you could cover some of the ML tools available on AWS such as SageMaker, it supposedly enables you to quickly and easily deploy a model and I've heard that it can be used in conjunction with TensorFlow.
Where do I find the download to the Keras file? I am trying to follow along and I'm stuck around 7:44
what editor are you using? to edit files.
hi i have a object detection model, using tensorflow object detection inception v2.
cant able to figure out output node name which is required for tensorflow js... any help...
Might be able to help, post some code in Slack paskuale.ibrushi@gmail.com
use parseInt(val) ....... as the angular takes the value as string rather than a number!
Can you site those coursera courses you'd take?
Does it have the full functionality like tensorflow with python ?
The API is about 90% covered. Feels similar in most ways, but there are some differences.
where the repo link for these two projects
nice post thanks...
i think gruvbox
I didnt realize it was this simple. FML.
How can I host my web application?
Hi is there any good courses about machine learning on udemy?
How is he pasting text into the editor like that?
Does anyone have any experience with real time object recognition or classification speed in tensorflow js vs tensorflow mobile? Which will be faster? I know mobile should be faster but I would need to learn Android development and I would prefer to know for sure before I start on either.
What VS code theme do you use?
Do you have the code available for the simple linear regression model?
Hey! I just uploaded a video on a regression video in Tensorflow! Check it out!😊👍
thanks for your video , could you make a video to verify if a mail is spam or not spam iin tensor flow ....... pls from nigria
That's a good idea. Spam filters were one of the first successful uses of neural nets.
we could actually use dataset on kaggle too.... will be very glad if u approve my request.... Kudos too
A video like that is highly likely - Kaggle is where I usually get my training data.
cool bro.. just recently got into ML and u just inspired me into thinking big .... much love bro
everything bounced over my head
1:46 ambition spotted
How can I provider multiple arrays as inputs? I have a CSV file which i want to import? I am really new here and its frustrating me that I can't find a simple anywhere.. :(
I want to have something like this:
[1,2,3,4,5,6], [7,6,5,4,3,2,1]]
then the corresponding output for the first set of numbers would be 10,20,30,40,50,60
and for the second it would be : 70, 60, 50, 40, 30, 20, 10
Can someone please help?
CZcams keeps deleting my replies with code example, don't know how to write it so it wouldn't be deleted :(
Call flatMap on that array with arrow function as first argument, name first argument of arrow function as "value" (for example) and multiple "value" by 10.
Okay! This is really important! What theme are you using for VS Code?
atom-one-dark + vscode-icons
man what is that music in the background at the start
please do EMOJI SCAVENGER HUNT in ionic :))
github.com/google/emoji-scavenger-hunt
Can you make a TensorFlow in C video?
To this day, are those resources still relevant to learn ML?
what theme is this in vscode
Tensorflow 100 seconds pleasee :>
This video is tensor tensor tensor flow tensor. Oh and tensor and tensor flow
SGD stands for Singapore $ please 😁
How to run the app?
nvm. I found it out. "ng build" then "ng serve" from the src/app
Demo link not working.
Thank you, just fixed that.
Is this js or ts? I see some grammar from ts. If tf works with ts, then I guess it's time for me to learn tf.
Video from the past. When there is no crypto, no a.i ,no layoffs bullshit was there. 😢😢
Man do the overall front-end code with React please. No one uses Angular so generally & I think it is the only reason you're getting better views but less watch-time, may be
I would prefer you to type the code rather than pasting it. :)
pfff, my brain is on fire
3:03 why does this take "machine learning" ? seems like a straightfoward math equation
This is awesome my man! Great job! I've been inspired to start my own journey to learning AI as well. I just published my first video! Check it out! 😊👍
I tried training a model with values of a function y = 2x and it returns totally unexpected results. What am I doing wrong? Here is the source code for node.js gist.github.com/LukaszWiktor/748c40c8a51514cc37f5e0e037becf1d
"Because you're most likely a human being"
Oh yeah, easy.
I didn’t have a background.
Dude, when did you learn all the stuff? Do you even sleep? 😅
For make a class in this way I'd rather not, very bad explication.
I just want to be honest. This quick start tutorial was too abstract. I need to be able to visualize the model and be able to know what each line does in the model.