I was inspired and learned the basics of TensorFlow after I completed the TensorFlow specialization on coursera. Personally I think these videos I created give a similar understanding but if you wanna check it out you can. Below you'll find both affiliate and non-affiliate links, the pricing for you is the same but a small commission goes back to the channel if you buy it through the affiliate link which helps me create more future videos. affiliate: bit.ly/3JyvdVK non-affiliate: bit.ly/3qtrK39
@aegoni61763 жыл бұрын
As someone who has watched all the tutorials videos, took notes and studied them to heart, and now making machine learning programs on my own, I really, REALLY appreciate all the efforts that you have done in your videos. Bless you man
@shivammishra73062 жыл бұрын
Took Notes of what, and studied how? (I am interested in this approach, thus asking this)
@aegoni61762 жыл бұрын
@@shivammishra7306 Written down the codes, then i started to analyse what each line of code meant to do. Over time and practice it will be an easy thing, but you have to try out the code of each video for yourself to see what it does, like the plt.scatter(x,y). Play around with the code and take not of each step you find important. Im kind of a nerd myself, so i wrote down everything each line of code did. It took some time, and maybe its not the best step to learn, but i highly encourage you to do something similar if you wanna learn strongly, or at least take a good idea of the basics
@hemonn Жыл бұрын
Thanks for your work!! I will study from Japan
@xavierrispal96053 жыл бұрын
Thanks for this tutorial series. Would you plan to cover one-class classifier aka oc-svm coupled with deep cnns used as feature extractors ? Best regards :)
@wolfisraging4 жыл бұрын
Thanks for the video, I wish these videos were there last year.. when I used to struggle doin these basic projects... :}
@AladdinPersson4 жыл бұрын
Thanks for the kind words bro, you working on any video rn? :)
@wolfisraging4 жыл бұрын
@@AladdinPersson yeah... its something related to gans... you'll see it soon :)
@deepakkumar-qg2xh3 жыл бұрын
This was very Helpful...
@mohammedkassab31103 жыл бұрын
Thank you so much, this was good.
@riis084 жыл бұрын
@Aladdin Persson thanks for your amazing videos. #1 videos in my ranking on tensorflow, already passing to friends but still you don't have much audience. I hope in future you will start some project with GAN, UNeT, transformer etc. Thanks for amazing contents.
@AladdinPersson4 жыл бұрын
I really appreciate your kind words 🙏 I'm still learning and growing my knowledge in many areas including how to make better videos through clearer explanations, better quality etc. How I view it is that as I make more and better content and people find the videos useful a larger audience will be there, but time will tell! With that said there are quite a lot of people following me already, 3000 people is a lot of people and I'm very grateful for those people and people like you that drive me forward
@user-or7ji5hv8y4 жыл бұрын
Would love to see such types of projects in PyTorch.
@muhammadnaveed30863 жыл бұрын
#Thanks for your great efforts. Love from Pakistan......
@Shivam98i3 жыл бұрын
Hey Aladdin, Great content ♥️ I got to learn a lot from you... Can you tell what's your pc specs?
@AladdinPersson3 жыл бұрын
OS: Dual boot on separate hard drives Windows (for making videos) and Ubuntu for coding CPU: i7 8700k RAM: 32 gb ddr4 GPU: gtx 1060 i need a new one :(
@azizbayramli79163 жыл бұрын
Hey Aladdin, thanks for the great tutorial and for sharing your knowledge with us. I tried your model by myself and I got a precision score for benign 0.82 which is good and for malignant I got 0.18. I tried to predict on single images and for benign images, I get a good result and unfortunately, malignant predictions are way off. How can I improve the model so I can get a good prediction for the malignant class?
@kaiye49543 жыл бұрын
Thanks for the video. Why in the following for loop to add if statement for check num_batches? When iterates test_gen in each for loop, it returns one batch of test_gen? for _, y in test_gen: test_labels = np.append(test_labels, y) num_batches += 1 if num_batches == math.ceil(test_examples / batch_size): break
@JY-er5iy Жыл бұрын
Appreciate this tutorial video. One question, when we use the ImageDataGenerator function, does this mean all our training images are augmented ? i.e. the model will not be trained on the original, un-augmented images?
@fajarputra71692 жыл бұрын
Thanks for the video sir. Your vid really help me in making a code for my undergraduate. I have a question, if I want to evaluate my model like ini this video but my model is a multi-class can it be done using a library or I have to do it manually? thank you
@Manveer_Dhindsa Жыл бұрын
I get this error: TypeError: Cannot interpret 'tf.float32' as a data type
@asmaeennaji7332 жыл бұрын
i can t dowload the data, my email is not valid, please help
@Sara-yy7yl2 жыл бұрын
when I type my email,appear these "error Something went wrong, try again later."
@SamA-xu3dz3 жыл бұрын
Hey Aladdin, thanks a lot for the series of tutorials. I have gone from knowing nothing about machine learning to writing my own algorithm in under a week. Much clearer and more concise than anything else iv found. I was just wondering, is there any way to run this using larger images? Mine are 3200x3200 and re sizing them removes a lot of important detail.
@AladdinPersson3 жыл бұрын
Really appreciate the kind words! You can run it on larger images (you might have to adapt the CNN you're using) but one con here is that it will require a lot of VRAM on your GPU, so 3200x3200 would perhaps be difficult but try it
@tsovinar86 Жыл бұрын
Sorry for asking but if numpy does not work on pycharm or vs code can we program all of it on jupyter notebook?
@apolanco32253 жыл бұрын
how many epochs did you use to have those results? thanks for the video!
@fremaxter3 жыл бұрын
Can you do a video on tf.data, I want to train a lstm with large data that doesn’t fit ram. Or just generally use a good input pipeline Thanks a lot, just recently found you, you’re doing a good job :)
@AladdinPersson3 жыл бұрын
I've done two videos on loading custom image and text data, and in those videos I've shown various ways where some of them includes using tf.data, maybe that would help?
@abhijeet64312 жыл бұрын
we want yolo, retinaNet, efficiendNet ,and many more algorithms from scratch in tensorflow 🥺 and also how to use tensorflow Hub and transfer learning and fine tunning in tensorflow 🙏 please
@maxmunch9252 жыл бұрын
Can someone please help with how I can test different models using the hub. And also how I can randomly shuffle images in a folder to train test val folders like here
@nataliaresende27413 жыл бұрын
Hi Aladdin, in a paper on NLP I was advised to 'to take the last hidden layer of the network and projecting it to 2D plane through a dimensionality reduction technique such as PCA, we can discover the general clusters of words" Do you how to do that?
@AladdinPersson3 жыл бұрын
I haven't implemented principal component analysis from scratch but you should be able to use sklearn: scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html They have a couple of examples they show at the bottom:)
@nataliaresende27413 жыл бұрын
@@AladdinPersson thank you so much! I'll check this out
@krzusiekniwazne4 жыл бұрын
Are you going to work with tf.js or will it always be python?
@AladdinPersson4 жыл бұрын
As of right now my knowledge in javascript is rather limited (nonexistant) so I would have to study up on that part before making videos on it. I do want to create websites utilizing deep learning so I will probably learn about it sooner or later but I guess the question is when
@mustafachenine79422 жыл бұрын
is there the same project using pca
@nalinnagar12323 жыл бұрын
Just asking, what framework do you recommend I should learn, PyTorch or TensorFlow?
@AladdinPersson3 жыл бұрын
What are your goals with deep learning?
@nalinnagar12323 жыл бұрын
My goals are to implement things like CycleGAN and Attention, I feel like TensorFlow is just as good as PyTorch now because of the introduction of TF 2.0 but what of you think?
@AladdinPersson3 жыл бұрын
@NALIN nagar If your goal is to learn about deep learning and try your own projects and not focus too much on production (making websites or apps) I would choose PyTorch. This might be a personal bias since I started with it, but I feel it has several advantages which I could go into more detail on but the essential is that when I code in PyTorch it feels like a natural extension of Python. I feel like I am coding in Python, with an added concepts of tensors, whereas coding in TensorFlow I feel is much more a framework/language on it's own that is not as connected to just "normal" Python. With that said, both are great and you can make both work for sure, and in the grand scheme of things this decision doesn't really matter. This is still my opinion after coding in TF 2 although it is clearly much much better than TF 1 :)
@nalinnagar12323 жыл бұрын
I already know TensorFlow but don’t know PyTorch so do you recommend I switch? Because I do want to learn and do my own projects?
@AladdinPersson3 жыл бұрын
@@nalinnagar1232 No, keep with TF :)
@jeelanibasha39845 ай бұрын
To the person who have seen all this videos,does all the concepts are covered?? please provide the suggetion whether to invest time or not plz reply anyone...
@miguelxplayer96415 ай бұрын
YES!!! I can now make a full AI
@IV__VishnuSasidharan2 жыл бұрын
Hi can any one plz tell me how to run this program in google colab
@chinmayakaundanya31513 жыл бұрын
is anyone getting an error while downloading the dataset?