TensorFlow Tutorial #01 Simple Linear Model

How to make a simple model in TensorFlow for recognizing handwritten digits from the MNIST data-set.

This tutorial has been updated to work with TensorFlow 2.1 and possibly later versions using "v.1 compatibility mode".

[Mar]

Brilliant explanations.  Within our Internet haystack upon haystacks of dung, it is refreshing to find valuable gems such that you'd provided.  Too bad we can't clone you, your tutorials, and your wisdom.  This world deeply needs knowledgeable professionals like yourself!  Bravo and cheerio, chap!!

[younisibrahim]

I challenge and bed anyone can find a series of tutorials about TF more detailed than this one on YouTube. Thanks a lot for your great effort.

[jjjj_]

Your voice is so easy to understand in your words stick to my brain. Also the pun/joke you include every now and then are always funny!!

I love your work!

[skieyangel]

Your tutorials are my latest favorite tutorials! I have checked out a few and love how efficient you are in these videos! Thanks and keep them coming!

[mindaugasm]

Really good and thorough tutorials. Most of the bits shown here I have already seen on the internet, but your value is the proper way of connecting all the concepts and explaining intuition behind some tougher concepts. Thank you.

[lpopuuw]

This is hands down the best tutorial series I have ever come across. Combined with jupyter it is simply awesome! Thank you very much.

[jianxiangfeng]

Very good and clear explanations! I feel regretful that I have not found these videos before!!!

[jaguar_akki]

I was curious about all this new Deep Learning stuff I was hearing about, so I tried looking at the online lectures and courses that were available. But I found them to be confusing and frustrating.
I am facing the same problem but finally I get you thank allot man amazing tutorial.

[ijyotir]

One of the best tutorials on TF available so far. The way in which you have given line by line explanation helps a lot.
I have gone though the first three videos and the git hub code, really appreciate the effort you have put .

In the first three videos I understood how the code works and loss is reduced with different algorithms, it will be nice if you give a hint how to use these trained models by passing a 28x28 or nxn size image from outside.

[BatShitAI]

Thank you very much for the clear and detailed tutorial. I have viewed tons of videos and taken a good deal of online courses (including paid ones) and this video along with the accompaning files are, as Vivek put it, PURE GOLD! I can't thank you enough. I am really looking forward to going through the whole series and trying out all the code and exercises. Best regards!

[jeremycrowe]

Thank you for this gentle introduction to Tensorflow.

[tehCRITIZIZER]

Thank you these tutorials. They are seriously good! Much better explanation of the library functionality than googles own. The attempt to dumb it down in their own intro-tutorials, is doing more harm than good imo.

[joannwang]

These series tutorials are best. Clear explanation, very useful! Thanks!

[jaystormvlogs]

Haven't even finished watching the video and I subscribed because your the first channel Ive come across that has awesome video tutorials on Tenserflow and Deeplearning.

Thank you, and please keep up the amazing videos.

[PALYGAP]

Bravo !!!

Definitely the best Tutorials I've seen for learning TensorFlow and Convolutional Neural Networks and ...

Thanks a lot for providing the corresponding JUPYTER(Ipython) Notebooks. It makes it so much easier to fiddle with (and eventually break) all these models :)

You're top ;)

[karkinissan]

"Maybe if you smoked some weed or something.." LMAO, that part made me crack up real good.

[gue]

I can only second and underline the two dozen compliments re your TensorFlow tutorial!
Thanks a ton!
G.

[aopt]

best tensorflow tutorial I've ever seen!

[fgfanta]

Neat! Modifying your working example is a great way to understand and learn Tensorflow. Thank you sir!

[JHauris]

OUTSTANDING!!! What a nice, clear explanation. Well done. and thank you, Jon