Dude ! You know what you are doing ? You are taking hard to access knowledge and making it available to people. You know what that makes you ? That makes you awesome. Keep it going man. I love watching your videos where you implement stuff and show the code.
I love what you're doing - bringing data science/machine learning to the masses. Don't get me wrong, I'm a CS major myself and as outgoing as they come but you have the most magnetic personality of all the CS RUclipsrs. You're like the Freddie Mercury of Computer Science...which is a poetically apt comparison since he was desi too.
I rarely comment on videos, but I have to say that your videos are amazing! Aspiring data scientist here, and your videos got me motivated to keep on studying these advanced concepts, please keep up the good work :D
I became part of Patreon for the first time just to support you, thank you Siraj keep it up! That said I think this is a good channel but I know you can do much better. I can not wait for the future!
Ok, I am a visual artist, and it all really is LEGIT ! We are AI "free self creative" machines, that "decide" what to do with the data that we are constantley recieving. And AI is doing the same, we just haven't given them the fully automacy of decision making. that apparently we have. Only then would it be possible for AI to be better than us.
for those who dont know theres also whats called "Deep Style" which the computer paints over any given picture with an influence from over a 2nd image... the results are crazy check it out
this is the first video of an Indian tech guy that I totally like! However I would argue on the possibility to make a real perception of the reality under mushrooms, because what you "see" when you are deep under the mushrooms is not all that comes through your eyes. Probably in a later future, when Elon Musk creates the interface to communicate with the limbic system we will be able to affect the perception and change deeply.
Most of this went right over my head, but it was still highly interesting! I'm leaving this comment in the hope that one day there will be a simple "get this app" response to enable neural style transfer to video for numpties like me... :)
MetalSombie Yes the learning curve is very steep and also requires you to be already proficient in tenserflow, otherwise brain gets clogged with so much information
To install TensorFlow without errors on windows check for your Python version : - Use a 3.5.* (check with >>> python -V in cmd) - Then excecute >>> pip3 install --upgrade tensorflow
I went to command prompt and input *_pip3 install --upgrade tensorflow_* and got a ton of errors. Then I tried *_pip install -U tensorflow_* and everything worked out fine. is there anything that you method achieves that mine didn't?
What was the effect which you did at 0:12 ? and can we do this as well as the effect given by winner of this challenge on live video? As in I am recording myself and the live output is trippy. The video of winner of this challenge did not have sound, so is there any issue in sound rendering?
its a pity there are several steps that are not explained , it would be added an overall explanation of the project , how the background image looks like , how the other image looks like , should we add our own image ? , how is the basic image treatment ? , etc , in order to be more insightfull ., thanks
Am I the only one who thinks that the complexity jumps between the videos are a bit far? :D I am going to try this with my own image... Not a video straightaway.
I understand the code and the concepts, but I am confused about the tools I should use to write my code. I have access to a GCP and writing code on a Python notebook, but what If I want to make a web app? Can you explain how I make my program available for every device? Maybe you already have a video. Thank you!
Hey Siraj, awesome video as always! Just one little question: What might be the reason for my output not being as creative as in the video? I can notice some changes but not much. Kind regards. :)
@Siraj Raval About your statement at roughly 1:30: It's actually the other way around. It doesn't matter how a system is done. A computer could in theory be made out of syrup. Or chemicals (google it) for the matter of fact. So it is entirely thinkable that a neural network built in TensorFlow and a neural network built in chemicals (our brain) get similar results with similar training. Even if the one is entirely abstract, it doesn't matter. The other way around it is entirely possibleto simulate parts of our brain in software.
Hi Siraj!!!! Keep up the good work, i am doing an Ai startup in the Art, w/ gallery pieces in mind but i'M also very interested in Street art, from wheat pasting Ai generated images and so on.....Not sure ok......You have a great channel for inspiration and motivation!!! Peace out!!!!!
Awesome dude, I have successfully ran the code on my machine...... Now I have some doubts regarding it 1) How actually we are maximizing the layer activation using gradient ascent algorithm.(Basic description please) 2) Why we are splitting image in octaves, where we are just resizing the image; I have tried it for without resizing and I am not getting desired output, My question is WHY ?
Hey guys, I am not sure about the first argument in the line 'render_deepdream(tf.square(T('mixed4c')), img0); i know tf.spuare returns the element wise square of the array, but the argument T('mixed4c'), what is that??
with open(local_zip_file,'wb') as output: FileNotFoundError: [Errno 2] No such file or directory: '../data\\inception5h.zip' could someone tel me what im doing wrong and how to fix it? siraj??
This video freaks me out because I spent about an hour to rewrite the code from the screen with no idea of what Im doing to figure out that there's more code to copy from github, and the code from video was updated so basically I had to just copy github code and had no idae on how this works. THIS IS NOT LEARNING, and Im sad for wasting time. But I belive you have more detailed videos, and I like you style, just not the useless superfication...
My goal is use Neural network and recognize each letter (eg: a,b...x,y,z, A..Z..0...9). Now,I can segment each letter from the picture and I learn tensorflow by running example from this tutorial github.com/nlintz/TensorFlow-Tutorials and reading this book deeplearningbook.org. So what I confusing now is how to find a good model or architecture for tensorflow to recognize letter. Can I use the same those MNIST classifiers (yann.lecun.com/exdb/mnist/) for recognizing English alphabet/letters ? Cause MNIST classifiers recognize digit (0 - 9) so I really don't know It could handle OCR or not ?
when I try ur codes in my PC,it gets error like this File "D:\Program Files\Anaconda3\lib\site-packages\zmq\utils\jsonapi.py", line 43, in dumps s = s.encode('unicode') UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 2203: surrogates not allowed what should I do?
Hey, Siraj I'm a newbie and need your help a little bit. when I'm trying to pip install the dependencies for few an error message is showing in the windows cmd like " failed building wheel for functools or could not find a version that satisfies the requirement.
i learned it via the internet. a bunch of smart people helped me learn it by posting tutorials, papers, and code on the Web. I read it consistently for months. www.reddit.com/r/machinelearning is a great source. also a few online courses. i did learn it at school as well, but the real learning came from the internet.
Hey Siraj, I have heard that the math is very complex for machine learning. Is this correct ? I am good learner but I don't think I am good at abstract math concepts. Wouldn't there be something simply to deploy like a library (a jQuery equivalent) which can do all the heavy lifting, rather than learn calculus ? I have heard a lot about TensorFlow, which seems to be like that, but haven't gotten into learning it yet.
For machine learning you will need multivariable calculus, probability/statistics and linear algebra, he made a video about it recently. In theory, you could start playing with some simple algorithms just like he does in these videos, but to know those subjects would be really important to undestand what the program is actually doing or even decide the better algorithm for each situation and data set
Windows 10 now offers Bash on Ubuntu, as of the Anniversary Edition. That means that you can get most of the non-graphical features of Bash inside Windows
I tried to see the video so many times its became to too fast and to understand than I decided to set the playback speed as 1.5x now I am understanding clearly but concept was easy need to learn more. Set 1.5x speed and try to hear 😊
Hey Siraj, I installed tensorflow on Ubuntu 16.04. But every time I try to import tensorflow, I get following issue. ImportError: libcublas.so.8.0: cannot open shared object file: No such file or directory Failed to load the native TensorFlow runtime. I tried tensorflow page for possible problems, but this one persists on my system. This is due to missing library, but hopefully this should have installed by the tensorflow. Any idea why this happens?
I'm throwing an error 'data_dir' is not defined... I tried copying the official siraj code from github and i got the same error.. anyone else hitting the same setback?
@@awesomefeldmanfamily I made a few changes, mainly by removing the part that downloads the zip, and adding a saved file functionality: github.com/aztecman/deep_dream_soai/
File "/Users/Jay/Desktop/Deep Dream.py", line 18, in main with open(local_zip_file, 'wb') as output: FileNotFoundError: [Errno 2] No such file or directory: '../data/inception5h.zip' I cant figure out how to fix this, ive looked at everything as hard as i can but cant see it xD
I figured this error out through brute force so i think thats covered xDD however, im now getting this error: File "/Users/Jay/Desktop/Deep Dream.py", line 78, in main render_deepdream(T(layer)[:,:,:,139], img0) NameError: name 'T' is not defined Despite the fact that ive copied your code practically exactly
OpenCV has many tutorials on how to do that. It doesn't use TensorFlow. Check out this out www.docs.opencv.org/master/d7/d8b/tutorial_py_face_detection.html
Hi thanks for your response.I would like to use Tensor flow and implement the face detection .so I would want to know how to use the pretrained classifier and detect the faces as it is done in other open CV projects . Do u have any idea on that ?
Yes. Check out the opencv website. They have tutorials on how to do it there. You can even make your own classifiers if you want, but they give you a really good face classifier with opencv.
Thanks Ryan.I have seen that Open Cv has classifiers which could do face detection.I would like to know how we could do it using Tensor Flow using Image Net's pretrained model or build using Deep ConvNets. Please let me know if you have any idea regarding that
Any help with this issue would be super great! I'm getting a syntax error on this line (img0 gets highlighted): render_deepdream(T(layer)[:,:,:,139] img0) I'm on Python 3.6.4 x64 bits. Windows 8.1.
ok im completely lost. Let assume i wanted to understand all of what you said in one year time. What should i do. I'm just a simple backend developer who builds crud applications. I understand I'll need to learn python, i know ruby so python looks easy. I asked someone they said first go learn/review Calculus, Linear Algebra, Statistics. True? I can do that in 5 months. Then where do i start. Someone told me to do Coursera course on machine learning. What about your channel and udacity courses. Where do they fit.
Dude ! You know what you are doing ? You are taking hard to access knowledge and making it available to people. You know what that makes you ? That makes you awesome. Keep it going man. I love watching your videos where you implement stuff and show the code.
thanks so much Ranjeeth! This is my life's calling. I will inspire and educate until my last breath
As long as he credit the researchers and doesn't claim to have found out all that himself, it's cool.
I love what you're doing - bringing data science/machine learning to the masses. Don't get me wrong, I'm a CS major myself and as outgoing as they come but you have the most magnetic personality of all the CS RUclipsrs. You're like the Freddie Mercury of Computer Science...which is a poetically apt comparison since he was desi too.
thanks so much Arannya. I've always loved Freddy Mercury
Freddie* :P
2:35 import numpy as numpy, well done ;)
LOL you caught me thanks
I don't usually comment, however ... this 'numpy as numpy' was, in deed, grand. hats off ;-)
This has VERY QUICKLY become my favorite youtube channel. You're doing a killer job.
I couldn't follow large parts of that, but basically, you are my new hero. I love what's emerging Deep Dream!
I rarely comment on videos, but I have to say that your videos are amazing!
Aspiring data scientist here, and your videos got me motivated to keep on studying these advanced concepts, please keep up the good work :D
thanks so much Luis :)
Your videos are really awesome, Siraj. It's much easier to learn Deep Learning and ML with your videos. Keep doing the great work!
This series is sick, this video was especially a treat!
I became part of Patreon for the first time just to support you, thank you Siraj keep it up! That said I think this is a good channel but I know you can do much better. I can not wait for the future!
thanks Diego!
Ok, I am a visual artist, and it all really is LEGIT ! We are AI "free self creative" machines, that "decide" what to do with the data that we are constantley recieving. And AI is doing the same, we just haven't given them the fully automacy of decision making. that apparently we have. Only then would it be possible for AI to be better than us.
After watching some of your videos....I BECOME A HUGE FAN OF YOUR SIRAJ RAVEL!!!
YOU ARE AWESOME!!!
awesome thx
for those who dont know theres also whats called "Deep Style"
which the computer paints over any given picture with an influence from over a 2nd image...
the results are crazy check it out
exactly thanks for saying that
Siraj Raval Hi! Could you make a video on how to make custom styles in tensorflow? Been trying to figure it out myself.
New Disciple. thank you siraj, great pace actually, and nice bumps of entertainment
thanks Andrew :)
i have just become a coding beliver.. great n fun teacher
wow! man, thanks for posting so informative videos in a really nice way!
I'm learning a lot!
all the best!
Thanks so much Cleyton
this is the first video of an Indian tech guy that I totally like!
However I would argue on the possibility to make a real perception of the reality under mushrooms, because what you "see" when you are deep under the mushrooms is not all that comes through your eyes. Probably in a later future, when Elon Musk creates the interface to communicate with the limbic system we will be able to affect the perception and change deeply.
Dude, your are just awesome, such a good job. You help many people, thanks!
thanks so muhc!
Most of this went right over my head, but it was still highly interesting! I'm leaving this comment in the hope that one day there will be a simple "get this app" response to enable neural style transfer to video for numpties like me... :)
You were born to do this.
Sirajology, can you make a video about yourself? What did you study, where do you live and so on
quebono100 nope.... lets keep some shroud of mystery shall we
hmm. i should. but it seems like Ramsey's comment got more likes. maybe people like the mystery?? lol
yeah, we love you SiraJology, keep on doing videos
It doesnt hvae more likes anymore
always with those fresh vids!
Great take on Deep Dream.
thanks Nick!
siraj .. you aren't good at all .. you are amazing and awesome bro 😻
thank you so much
Awesome videos..! Thank you siraj ☺
thanks!
The First two minutes were amazingly interesting and theeeeeeen it got intelligent and I stopped understanding anything...
Give a try once again.
MetalSombie Yes the learning curve is very steep and also requires you to be already proficient in tenserflow, otherwise brain gets clogged with so much information
MetalSombie fake it till u make it. there is no other way!
i feel like these are basically speed-throughs to show you how to do some cool thing IF you already know TF, NN's, Python, etc..
hahahahahah xdd
Great work Siraj.
thank you!
Man, you're awesome. Keep it up the good work! :)
If the doors of perception were cleansed everything would appear to man as it is, infinite - Aldous Huxley
To install TensorFlow without errors on windows check for your Python version :
- Use a 3.5.* (check with >>> python -V in cmd)
- Then excecute >>> pip3 install --upgrade tensorflow
I went to command prompt and input *_pip3 install --upgrade tensorflow_* and got a ton of errors.
Then I tried *_pip install -U tensorflow_* and everything worked out fine.
is there anything that you method achieves that mine didn't?
Siraj Bhai, Badhiya banate Ho!! Ek din saath me kaam karenge :)
thanks Tushar bhai
we love you siraaj u are going to be huge m waiting for that day
The new music video for Foster The People's song called, "Doing It for the Money" uses Deep Dream and it's trippy. Pretty sweet!
What's he doing? ... He is beginning to believe.
it starts
What was the effect which you did at 0:12 ? and can we do this as well as the effect given by winner of this challenge on live video? As in I am recording myself and the live output is trippy. The video of winner of this challenge did not have sound, so is there any issue in sound rendering?
play.google.com/store/apps/details?id=com.smaper.artisto&hl=en
+Sirajology and what about the live filter of deep dream?
its a pity there are several steps that are not explained , it would be added an overall explanation of the project , how the background image looks like , how the other image looks like , should we add our own image ? , how is the basic image treatment ? , etc , in order to be more insightfull ., thanks
Am I the only one who thinks that the complexity jumps between the videos are a bit far? :D I am going to try this with my own image... Not a video straightaway.
I understand the code and the concepts, but I am confused about the tools I should use to write my code. I have access to a GCP and writing code on a Python notebook, but what If I want to make a web app? Can you explain how I make my program available for every device? Maybe you already have a video. Thank you!
Hey Siraj, awesome video as always! Just one little question: What might be the reason for my output not being as creative as in the video? I can notice some changes but not much. Kind regards. :)
@Siraj Raval About your statement at roughly 1:30: It's actually the other way around. It doesn't matter how a system is done. A computer could in theory be made out of syrup. Or chemicals (google it) for the matter of fact. So it is entirely thinkable that a neural network built in TensorFlow and a neural network built in chemicals (our brain) get similar results with similar training. Even if the one is entirely abstract, it doesn't matter. The other way around it is entirely possibleto simulate parts of our brain in software.
I'm poor at math.. but I'm pretty much interested in learning Machine learning. What are the must concepts to succeed in Machine learning?
This is amazing.
Thank You
Xx
Hi Siraj!!!! Keep up the good work, i am doing an Ai startup in the Art, w/ gallery pieces in mind but i'M also very interested in Street art, from wheat pasting Ai generated images and so on.....Not sure ok......You have a great channel for inspiration and motivation!!! Peace out!!!!!
What's the source on that visualization at 1:53?
Awesome dude, I have successfully ran the code on my machine...... Now I have some doubts regarding it
1) How actually we are maximizing the layer activation using gradient ascent algorithm.(Basic description please)
2) Why we are splitting image in octaves, where we are just resizing the image; I have tried it for without resizing and I am not getting desired output, My question is WHY ?
www.alanzucconi.com/2016/05/25/generating-deep-dreams/
Hey guys,
I am not sure about the first argument in the line 'render_deepdream(tf.square(T('mixed4c')), img0);
i know tf.spuare returns the element wise square of the array, but the argument T('mixed4c'), what is that??
Great videos ! keep them coming. How about a video on AI based text summarization ?
thanks! i like your idea, i've written it down
with open(local_zip_file,'wb') as output:
FileNotFoundError: [Errno 2] No such file or directory: '../data\\inception5h.zip'
could someone tel me what im doing wrong and how to fix it?
siraj??
I have the same error??
why isnt the data directory being created
This video freaks me out because I spent about an hour to rewrite the code from the screen with no idea of what Im doing to figure out that there's more code to copy from github, and the code from video was updated so basically I had to just copy github code and had no idae on how this works. THIS IS NOT LEARNING, and Im sad for wasting time. But I belive you have more detailed videos, and I like you style, just not the useless superfication...
How do you use your own image as the deep style? Great work, thank you for making information like this more accessible. Not all hero's wear capes.
Please do a more in depth, slower programming for ML, AI and tensorFlow possibly exploring more into tensorflow, i think it would be amazing
will do thanks Joseph
Is there any way to pass the layer and channel as argument in the code. Currently not seeing how to pass the channel as argument in the github code.
You're the perfect human
thanks Finlay!
thank you for tensorflow video
anytime
Hey Siraj, how did you do the mushroom scene in the beginning. That looks really cool
its an app called artisto. thanks!
Can you expand on the octaves? What would happen if I choose 16 octaves vs if I choose 2 octaves? What about iter_n?
thank you. I still learning how to OCR by using tensorflow. Do you have any thesis or documents, graph, model about it.
more vids on this soon
My goal is use Neural network and recognize each letter (eg: a,b...x,y,z, A..Z..0...9). Now,I can segment each letter from the picture and I learn tensorflow by running example from this tutorial github.com/nlintz/TensorFlow-Tutorials and reading this book deeplearningbook.org. So what I confusing now is how to find a good model or architecture for tensorflow to recognize letter.
Can I use the same those MNIST classifiers (yann.lecun.com/exdb/mnist/) for recognizing English alphabet/letters ?
Cause MNIST classifiers recognize digit (0 - 9) so I really don't know It could handle OCR or not ?
I dont know if you have done it or not but I have completed handwitten letters and numbers OCR reognition. Let me know i you need any help
Great video!
I give up,, that was supposed to be a beginner course.
Hi Siraj, you should teach how to make this with our own model
will do
I want to use a pic of my own. How can I insert it into this code?
This seems to be using TF 1. Could you update the video to use TF 2?
I'm tripping acid while watching this whaaaaaaaa
Damn, I love these videos but I realized i'm too dumb to get all of this
when I try ur codes in my PC,it gets error like this
File "D:\Program Files\Anaconda3\lib\site-packages\zmq\utils\jsonapi.py", line 43, in dumps
s = s.encode('unicode')
UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 2203: surrogates not allowed
what should I do?
That is amazing!!! Btw how could I retrain this model with other images?
NameError: name 'img_noise' is not defined
how did you fix this
@@JapiSandhu remove =img_noise.
@@JapiSandhu Full code: github.com/jnirschl/cell_decoder/blob/d9f00f954c0d3f0a807c3090e1d99dece1a3ecdc/cell_decoder/visualize/deep_dream.py#L53
Hey, Siraj I'm a newbie and need your help a little bit.
when I'm trying to pip install the dependencies for few an error message is showing in the windows cmd like " failed building wheel for functools or could not find a version that satisfies the requirement.
See the error log carefully, if you add VC++ for your python version it should solve your issue. I also had a similar issue.
@Sirajology How did you learned all this? By yourself or you learned it at school?
i learned it via the internet. a bunch of smart people helped me learn it by posting tutorials, papers, and code on the Web. I read it consistently for months. www.reddit.com/r/machinelearning is a great source. also a few online courses. i did learn it at school as well, but the real learning came from the internet.
Sirajology awesome thank you! :)
how do I know which video is the next one tho. I just want to know if anyone created the code for video
am I the only one getting a error on the tabs an indentions at the graph=tf.Graph() part it wont let me run it inside the def main() function
Hey Siraj, I have heard that the math is very complex for machine learning. Is this correct ? I am good learner but I don't think I am good at abstract math concepts. Wouldn't there be something simply to deploy like a library (a jQuery equivalent) which can do all the heavy lifting, rather than learn calculus ? I have heard a lot about TensorFlow, which seems to be like that, but haven't gotten into learning it yet.
For machine learning you will need multivariable calculus, probability/statistics and linear algebra, he made a video about it recently. In theory, you could start playing with some simple algorithms just like he does in these videos, but to know those subjects would be really important to undestand what the program is actually doing or even decide the better algorithm for each situation and data set
urllib.error.URLError: How can I fix this? I get this error when I try to run the script for the first time in the terminal (time=5.13)
you are my guru
thanks Ramsey :) i'll keep them coming for you
Hey Siraj - is there a windows based console that replicates everything you're doing?
Windows 10 now offers Bash on Ubuntu, as of the Anniversary Edition. That means that you can get most of the non-graphical features of Bash inside Windows
except for the actual packages
what that guy said
I tried to see the video so many times its became to too fast and to understand than I decided to set the playback speed as 1.5x now I am understanding clearly but concept was easy need to learn more. Set 1.5x speed and try to hear 😊
Hey Siraj, I installed tensorflow on Ubuntu 16.04. But every time I try to import tensorflow, I get following issue.
ImportError: libcublas.so.8.0: cannot open shared object file: No such file or directory
Failed to load the native TensorFlow runtime.
I tried tensorflow page for possible problems, but this one persists on my system. This is due to missing library, but hopefully this should have installed by the tensorflow.
Any idea why this happens?
Yes I had.
I'm throwing an error 'data_dir' is not defined... I tried copying the official siraj code from github and i got the same error.. anyone else hitting the same setback?
just create the folder that he is using in path, or change the path according to your needs
nope. this didn't fix it.
@@AZTECMAN help! I'm having the same error! Watt do I doooo???!!!
@@awesomefeldmanfamily I made a few changes, mainly by removing the part that downloads the zip, and adding a saved file functionality: github.com/aztecman/deep_dream_soai/
If you want I can explain exactly what I changed to make it work, but long story short, here it is!
File "/Users/Jay/Desktop/Deep Dream.py", line 18, in main
with open(local_zip_file, 'wb') as output:
FileNotFoundError: [Errno 2] No such file or directory: '../data/inception5h.zip'
I cant figure out how to fix this, ive looked at everything as hard as i can but cant see it xD
I figured this error out through brute force so i think thats covered xDD however, im now getting this error:
File "/Users/Jay/Desktop/Deep Dream.py", line 78, in main
render_deepdream(T(layer)[:,:,:,139], img0)
NameError: name 'T' is not defined
Despite the fact that ive copied your code practically exactly
post an issue in the repo
Sirajology okay, I managed to fix it on my own haha
hey dude, how did you fix the first issue [Errno 2] No such file or directory: '../data/inception5h.zip'?
Adam -亚当- I kinda just brute forced the location. hard to explain
you are a genius!
thanks so much!!
Hi,
Can we use the pretrained model of Image net and extract a layer and perfrom face detection?
If so How can we go forward with it?
OpenCV has many tutorials on how to do that. It doesn't use TensorFlow. Check out this out www.docs.opencv.org/master/d7/d8b/tutorial_py_face_detection.html
make a haar classifier that is stored as an xml and use that as a pretrained model that you load to detect faces
Hi thanks for your response.I would like to use Tensor flow and implement the face detection .so I would want to know how to use the pretrained classifier and detect the faces as it is done in other open CV projects .
Do u have any idea on that ?
Yes. Check out the opencv website. They have tutorials on how to do it there. You can even make your own classifiers if you want, but they give you a really good face classifier with opencv.
Thanks Ryan.I have seen that Open Cv has classifiers which could do face detection.I would like to know how we could do it using Tensor Flow using Image Net's pretrained model or build using Deep ConvNets.
Please let me know if you have any idea regarding that
6:21 why r ur eyes so wide here chill
I wanted to start machine learning from scratch..i am bit confused..how to start..which source ..plz help me out
see my latest video
Siraj your fkin amazing, keep up the style
Thanks Guy!
i so want these tutorials to work but they never do ..
neural nets exist on the circuit no?
Hey guys, is it possible to save the image that is generated to a specific resolution? pls help
Any help with this issue would be super great!
I'm getting a syntax error on this line (img0 gets highlighted):
render_deepdream(T(layer)[:,:,:,139] img0)
I'm on Python 3.6.4 x64 bits. Windows 8.1.
use a , before img0
Hello siraj,been wondering what is the link between data science and cognitive machine learning
do comment as early i have lot to discuss with you
machine learning is a tool for data science
Hello there!
been in love with Rasberry Pi,can you host a tutorial on the same.
it would be great then.
what program do you use to edit your videos?
final cut pro
+Sirajology thanks dude! ! you are awesome
Which text editor u recommend?
sublime
Who honestly could find out what he was saying at very first time watching the vid?
ok im completely lost. Let assume i wanted to understand all of what you said in one year time. What should i do.
I'm just a simple backend developer who builds crud applications.
I understand I'll need to learn python, i know ruby so python looks easy.
I asked someone they said first go learn/review Calculus, Linear Algebra, Statistics. True?
I can do that in 5 months. Then where do i start.
Someone told me to do Coursera course on machine learning.
What about your channel and udacity courses. Where do they fit.
watch my intro to deep learning palylist
Mhh it works but the generated pictures are kind of washed out. kind of like they only had a transparency of 80%.
Which text editor??i m bit confused
sublime
that's an unusual shirt
how to initialize t_obj
How old are you? Just curious
25
I'm fucked at very first step setting up environment. can't install dists
not fucked anymore
This tutorial no longer works with the tensorflow 2.0
What ide do you use
yup