Learn PyTorch for deep learning in a day. Literally.

Welcome to the most beginner-friendly place on the internet to learn PyTorch for deep learning.
All code on GitHub - dbourke.link/pt-github
Ask a question - dbourke.link/pt-github-discus...
Read the course materials online - learnpytorch.io
Sign up for the full course on Zero to Mastery (20+ hours more video) - dbourke.link/ZTMPyTorch
Below are the timestamps/outline of the video. The video you're watching is comprised of 162 smaller videos but KZbin limits timestamps at 100 so some have been left out.
00:00 Hello :)
🛠 Chapter 0: PyTorch Fundamentals
01:17 0. Welcome and "what is deep learning?"
07:13 1. Why use machine/deep learning?
10:47 2. The number one rule of ML
16:27 3. Machine learning vs deep learning
22:34 4. Anatomy of neural networks
31:56 5. Different learning paradigms
36:28 6. What can deep learning be used for?
42:50 7. What is/why PyTorch?
53:05 8. What are tensors?
57:24 9. Outline
1:03:28 10. How to (and how not to) approach this course
1:08:37 11. Important resources
1:14:00 12. Getting setup
1:21:40 13. Introduction to tensors
1:35:07 14. Creating tensors
1:53:33 17. Tensor datatypes
2:02:58 18. Tensor attributes (information about tensors)
2:11:22 19. Manipulating tensors
2:17:22 20. Matrix multiplication
2:47:50 23. Finding the min, max, mean and sum
2:57:20 25. Reshaping, viewing and stacking
3:11:03 26. Squeezing, unsqueezing and permuting
3:23:00 27. Selecting data (indexing)
3:32:33 28. PyTorch and NumPy
3:41:42 29. Reproducibility
3:52:30 30. Accessing a GPU
4:04:21 31. Setting up device agnostic code
🗺 Chapter 1: PyTorch Workflow
4:16:59 33. Introduction to PyTorch Workflow
4:19:46 34. Getting setup
4:27:02 35. Creating a dataset with linear regression
4:36:44 36. Creating training and test sets (the most important concept in ML)
4:52:50 38. Creating our first PyTorch model
5:13:13 40. Discussing important model building classes
5:19:41 41. Checking out the internals of our model
5:29:33 42. Making predictions with our model
5:40:47 43. Training a model with PyTorch (intuition building)
5:49:03 44. Setting up a loss function and optimizer
6:01:56 45. PyTorch training loop intuition
6:39:37 48. Running our training loop epoch by epoch
6:49:03 49. Writing testing loop code
7:15:25 51. Saving/loading a model
7:44:00 54. Putting everything together
🤨 Chapter 2: Neural Network Classification
8:31:32 60. Introduction to machine learning classification
8:41:14 61. Classification input and outputs
8:50:22 62. Architecture of a classification neural network
9:09:13 64. Turing our data into tensors
9:25:30 66. Coding a neural network for classification data
9:43:27 68. Using torch.nn.Sequential
9:56:45 69. Loss, optimizer and evaluation functions for classification
10:11:37 70. From model logits to prediction probabilities to prediction labels
10:27:45 71. Train and test loops
10:57:27 73. Discussing options to improve a model
11:27:24 76. Creating a straight line dataset
11:45:34 78. Evaluating our model's predictions
11:50:58 79. The missing piece: non-linearity
12:42:04 84. Putting it all together with a multiclass problem
13:23:41 88. Troubleshooting a mutli-class model
😎 Chapter 3: Computer Vision
14:00:20 92. Introduction to computer vision
14:12:08 93. Computer vision input and outputs
14:22:18 94. What is a convolutional neural network?
14:27:21 95. TorchVision
14:36:42 96. Getting a computer vision dataset
15:01:06 98. Mini-batches
15:08:24 99. Creating DataLoaders
15:51:33 103. Training and testing loops for batched data
16:25:59 105. Running experiments on the GPU
16:29:46 106. Creating a model with non-linear functions
16:41:55 108. Creating a train/test loop
17:13:04 112. Convolutional neural networks (overview)
17:21:29 113. Coding a CNN
17:41:18 114. Breaking down nn.Conv2d/nn.MaxPool2d
18:28:34 118. Training our first CNN
18:43:54 120. Making predictions on random test samples
18:55:33 121. Plotting our best model predictions
19:19:06 123. Evaluating model predictions with a confusion matrix
🗃 Chapter 4: Custom Datasets
19:43:37 126. Introduction to custom datasets
19:59:26 128. Downloading a custom dataset of pizza, steak and sushi images
20:13:31 129. Becoming one with the data
20:38:43 132. Turning images into tensors
21:15:48 136. Creating image DataLoaders
21:24:52 137. Creating a custom dataset class (overview)
21:42:01 139. Writing a custom dataset class from scratch
22:21:22 142. Turning custom datasets into DataLoaders
22:28:22 143. Data augmentation
22:42:46 144. Building a baseline model
23:10:39 147. Getting a summary of our model with torchinfo
23:17:18 148. Creating training and testing loop functions
23:50:31 151. Plotting model 0 loss curves
23:59:34 152. Overfitting and underfitting
24:32:03 155. Plotting model 1 loss curves
24:35:25 156. Plotting all the loss curves
24:46:22 157. Predicting on custom data
#pytorch #machinelearning #deeplearning

[@mrdbourke]

Who's ready to code some fire???? 🔥 Pun intended ;) Welcome to the world of machine learning my friend. It's a fun place. --- PS Don't forget to take breaks. Practice, break, practice, break. Enjoy both. Because much of learning happens when you're not doing anything, walking around, or taking a nap. And two ideas in your head collide. As for everything not mentioned in this video or learnpytorch.io, you'll find it in the PyTorch documentation (we'll reference it A LOT), because despite the length/title of this video, it's impossible to learn all of PyTorch in a day. Consider this video a momentum builder.

[@masternobody1896]

Epic

[@niranjannair390]

I hope you make more like this!

[@26_dhruvitmistry65]

I am ready here after doing your zero to mastery tensorflow course hope will learn a lot here! You are one of the finest teacher Daniel.

[@acnice_]

I'm here!

[@chenzihao6142]

🔥🔥🔥🔥

[@Dim-zt5ei]

Thank you, Daniel. I completed the 25 hours. I'm happy to see people putting in such great efforts to make knowledge accessible for free

[@mohammadashraf7420]

Thank you

[@gg_505_]

We live in a time where you can have this level of information for free. You are awesome dude, thanks for putting this up! I'll comment again once I finish the course.

[@mrdbourke]

You’re welcome G! Thank you for the kind words, enjoy :)

[@yuosef4642]

You finished the course?? I'm already 5 hours into it in two days

[@pintorafael]

I watched and clicked on every ad. Hope it helps him a little and thanks him for this amazing free information

[@SGworldview]

@@pintorafael that is just beautiful

[@sakthir2617]

Did you finished ?

[@teidenzero]

Day 4, I'm halfway through and I'm looking at code that looked alien to me 5 days ago and it starts to make sense. Absolute beast of a course.

[@dormansutt8429]

A few people said they didn't like that you keep repeating things, but I really appreciate you doing that because for deep learning is a very complicated subject, so you repeating stuff really helps me a lot. btw thanks for the course, Daniel

[@brycecolwell4304]

indeed, and i think most of us work that way or the majority of videos wouldnt be that way. My problem is sometimes they repeat the very simple and breeze over the ones that really need it.

[@mostafaalkady6556]

This gentleman was generous with his time and effort to teach people for free! I'm following the course now and it's very easy and I'm enjoying every bit of it. We need more people like you in the world!

[@mrdbourke]

Thank you Mostafa! I really appreciate it 🙏

[@skandhaprasath2767]

​@@mrdbourkeYou guys are starting a revolution. Please don't stop. This is the next gen. And this is the meaning of internet. This level of knowledge for free will absolutely shoot up the level of humankind in less than 100-200 years. It's gonna be a big brain revolution. Hope we can get materials like this in all domains. So that everyone can access them from everywhere, Not just the lads who has the time, coaching, money and support to get into Ivy leagues and learn these before their 20s.

[@mostafaalkady6556]

@@skandhaprasath2767 Agreed. Thousands of talents all over the world don't have enough resources to get degrees from schools, let alone universities. Making knowledge accessible is one of the noblest endeavors one can achieve.

[@ugestacoolie5998]

@@skandhaprasath2767 Im glad knowledge like this is just out there

[@muzammilomarzoy6616]

Day 1 | 31:56 5. Different learning paradigms Day 2 | 1:21:40 13. Introduction to tensors Day 3 | 1:53:33 17. Tensor datatypes Day 4 | 2:17:22 20. Matrix multiplication Day 5 | 3:11:03 26. Squeezing, unsqueezing and permuting Day 6 | 4:16:59 33. Introduction to PyTorch Workflow

[@chydra6538]

bro where r u now, cmon we need to make progress

[@30DaysMonkMode-ft1kf]

​@@chydra6538forget about this lazy bum bro, let us move on

[@dBanzy]

What a time to be alive 25 hours formation on PyTorch, I didn't even care about deep learning but now I want to create things with this new-found knowledge

[@mrdbourke]

Enjoy legend!

[@deepaksingh9318]

Totally loving it . loved the content, the simplicity the course is designed , and the way the course content is briefed is just perfect.

[@allcoolandnew]

Hi Daniel, wanted to make sure to thank you for sharing these great videos. You've been a big part of my daily routine for the last couple of weeks :) I feel that the fact that you've learned many of these things relatively recently yourself gives you a real edge because you still have fresh in mind the concepts you might've struggled with yourself. And your teaching style of repetition and positivity (I'm thinking: "Wow my internet connection is slow, but that's OK, it gives us more time to think about what we're learning here" 😄) really make the whole process fun and approachable. You're a great teacher! Expecting to see you in the bestsellers lists soon ;) Grts from the Netherlands

[@amandachang247]

This is amazing and honestly addicting to watch/practice with. Thank you so much for all the effort you've clearly put into this; I LOVE this tutorial and you're a great teacher.

[@mnh7813]

Finally completed the course!! Thanks for your work Daniel! It was a really fun journey to learn deep learning with Pytorch

[@mswiseman]

This is beyond incredible. I really appreciate you breaking everything down and providing outside resources.

[@farisrustom5194]

Fantastic video Daniel! Just finished the video after ~2 weeks and I can honestly say this is the best free educational resource I have come across, thank you so much!

[@TheLegend-mu6zg]

Finished the video after a couple weeks, doing around an hour of the video per day. Seriously this was great and was exactly what I was looking for. big thanks!

[@fr4864]

One of the best ML/PyTorch videos I’ve ever come across. Straight forward and packed with need to know information. We need more people like this in tech!

[@yashjindal9822]

Excited to complete this all the way! Thank you for this amazing course Daniel Will come back when I finish all the modules here

[@kostaskyr]

That's a phenomenal resource. Thank you for providing it to us. You are doing so much good for the community. Keep up the excellent work!

[@mrdbourke]

You’re welcome legend! Enjoy!

[@amortalbeing]

🛠 Chapter 0: PyTorch Fundamentals 01:17 0. Welcome and "what is deep learning?" 07:13 1. Why use machine/deep learning? 10:47 2. The number one rule of ML 16:27 3. Machine learning vs deep learning 22:34 4. Anatomy of neural networks 31:56 5. Different learning paradigms 36:28 6. What can deep learning be used for? 42:50 7. What is/why PyTorch? 53:05 8. What are tensors? 57:24 9. Outline 1:03:28 10. How to (and how not to) approach this course 1:08:37 11. Important resources 1:14:00 12. Getting setup 1:21:40 13. Introduction to tensors 1:35:07 14. Creating tensors 1:53:33 17. Tensor datatypes 2:02:58 18. Tensor attributes (information about tensors) 2:11:22 19. Manipulating tensors 2:17:22 20. Matrix multiplication 2:47:50 23. Finding the min, max, mean and sum 2:57:20 25. Reshaping, viewing and stacking 3:11:03 26. Squeezing, unsqueezing and permuting 3:23:00 27. Selecting data (indexing) 3:32:33 28. PyTorch and NumPy 3:41:42 29. Reproducibility 3:52:30 30. Accessing a GPU 4:04:21 31. Setting up device agnostic code 🗺 Chapter 1: PyTorch Workflow 4:16:59 33. Introduction to PyTorch Workflow 4:19:46 34. Getting setup 4:27:02 35. Creating a dataset with linear regression 4:36:44 36. Creating training and test sets (the most important concept in ML) 4:52:50 38. Creating our first PyTorch model 5:13:13 40. Discussing important model building classes 5:19:41 41. Checking out the internals of our model 5:29:33 42. Making predictions with our model 5:40:47 43. Training a model with PyTorch (intuition building) 5:49:03 44. Setting up a loss function and optimizer 6:01:56 45. PyTorch training loop intuition 6:39:37 48. Running our training loop epoch by epoch 6:49:03 49. Writing testing loop code 7:15:25 51. Saving/loading a model 7:44:00 54. Putting everything together 🤨 Chapter 2: Neural Network Classification 8:31:32 60. Introduction to machine learning classification 8:41:14 61. Classification input and outputs 8:50:22 62. Architecture of a classification neural network 9:09:13 64. Turing our data into tensors 9:25:30 66. Coding a neural network for classification data 9:43:27 68. Using torch.nn.Sequential 9:56:45 69. Loss, optimizer and evaluation functions for classification 10:11:37 70. From model logits to prediction probabilities to prediction labels 10:27:45 71. Train and test loops 10:57:27 73. Discussing options to improve a model 11:27:24 76. Creating a straight line dataset 11:45:34 78. Evaluating our model's predictions 11:50:58 79. The missing piece: non-linearity 12:42:04 84. Putting it all together with a multiclass problem 13:23:41 88. Troubleshooting a mutli-class model 😎 Chapter 3: Computer Vision 14:00:20 92. Introduction to computer vision 14:12:08 93. Computer vision input and outputs 14:22:18 94. What is a convolutional neural network? 14:27:21 95. TorchVision 14:36:42 96. Getting a computer vision dataset 15:01:06 98. Mini-batches 15:08:24 99. Creating DataLoaders 15:51:33 103. Training and testing loops for batched data 16:25:59 105. Running experiments on the GPU 16:29:46 106. Creating a model with non-linear functions 16:41:55 108. Creating a train/test loop 17:13:04 112. Convolutional neural networks (overview) 17:21:29 113. Coding a CNN 17:41:18 114. Breaking down nn.Conv2d/nn.MaxPool2d 18:28:34 118. Training our first CNN 18:43:54 120. Making predictions on random test samples 18:55:33 121. Plotting our best model predictions 19:19:06 123. Evaluating model predictions with a confusion matrix 🗃 Chapter 4: Custom Datasets 19:43:37 126. Introduction to custom datasets 19:59:26 128. Downloading a custom dataset of pizza, steak and sushi images 20:13:31 129. Becoming one with the data 20:38:43 132. Turning images into tensors 21:15:48 136. Creating image DataLoaders 21:24:52 137. Creating a custom dataset class (overview) 21:42:01 139. Writing a custom dataset class from scratch 22:21:22 142. Turning custom datasets into DataLoaders 22:28:22 143. Data augmentation 22:42:46 144. Building a baseline model 23:10:39 147. Getting a summary of our model with torchinfo 23:17:18 148. Creating training and testing loop functions 23:50:31 151. Plotting model 0 loss curves 23:59:34 152. Overfitting and underfitting 24:32:03 155. Plotting model 1 loss curves 24:35:25 156. Plotting all the loss curves 24:46:22 157. Predicting on custom data yes its available in the description, but its for ease of use ! Thanks mrdbourke

[@mgwyndolin4262]

Great content great presentation! You have the rare and amazing ability to arrive at the very essence of things before explaining them with an extreme clarity. I can never thank you enough for all the effort put into making this.

[@adamvo99]

Finished it all! Daniel, you are a true legend. Your teaching style is awesome and your enthusiasm is addictive. Thank you so much for helping so many people get started on their ML journey.

[@r0f115L4m]

You’re the greatest of all time, Daniel - thank you for putting this together!

[@mrdbourke]

You’re welcome Rofi! Enjoy :)

[@norf8]

This course is brilliant. I love the way how clearly you describe things. The best part is the number of sources you give as 'extra curriculum' - very very helpful. That being said, as a final year engineering student, I refuse to believe there are people who can be this happy in their lives. This shouldn't be legal. I don't know what's your secret but I need some of that energy :')

[@elitecoder955]

Just got started with machine learning and I have set a data on my calendar 2 months from now to come back to this video and learn PyTorch ! This channel is a goldmine

[@abhishekshinde9209]

Fun-to-learn award for all Machine learning courses goes to you !!! your courses are really really fun and interesting to watch. there is no point in your courses that i am bored !! Please continue your legendary work. i hope you make some more learning videos, i finished ZTM ML, ZTM tensorflow, and this now waiting for the next one

[@adnannoorullah5168]

This is literally amazing! Thank you Daniel! Super excited to learn more about Deep Learning. As a senior in CS, I can tell you that this style of teaching is awesome!

[@mrdbourke]

Thank you Adnan! Enjoy it legend!

[@starshipx1282]

You're a natural. I previously did the full tensorflow course and this is just another "blessing" of yours if I may say. Will be going with the full course on zm.

[@mrdbourke]

Thank you! If you enjoyed the TensorFlow course, you’ll love this one :)

[@user-tu7dn8ix4c]

OMG, I've just finished it! I'm so thankful for that, Daniel! I was so engrossed in process that it took me 4 days (~8 hours a day) to complete it. You are the best:*

[@wissammetawee]

Thanks so much Daniel for this great course, im a huge fan of the way you teach by repeating and explaining every little bit of code in details, and im like wow this guy is really a great teacher, actually after 13h of watching and writing everything you teach on my notebooks, which how i normaly study, i started to sing the loop song hehehehe, mate you are amazing , i really appreciate everything you are doing to help people like me who are struggling to acquire new knowledge.

[@shifty89]

One of the best (and the longest) instructional videos I've ever seen (still working through it). Thanks for this tremendous, free effort.

[@imoh5202]

A master legend at teaching and sharing. First, I love the fact that you stated what to expect before the beginning of each module and also structured the course in an easy to follow format with "Practical examples". This is what makes a great teacher. Second, is that the voice is steady in pace. This matters a lot especially for long tutorials like this. Once again thank you for your kindness and Cheer! Of course I subscribed immediately. First time here❤️

[@mrdbourke]

Wow! Thank you so much for the kind words, those are some massive compliments. Welcome to the channel :)

[@ting9252]

You are a very good lecturer! Easy to follow, clear step-by-step, real beginner applicable videos!👍

[@xiongjiederek]

Best deep learning online course I have seen so far. Thank you!

[@adriantsang311]

I just finished going through the entire video and I am so grateful for all the time and effort you have spent to put together this very organized and educational course to learn PyTorch for a beginner. Very well done!

[@mrdbourke]

Glad you enjoyed it Adrian! Thank you for the kind words!

[@tsukuruuu]

Day 1: 3:23:01 Nov 6 2023 Day 2: exercise + 5:13:12 Nov 7 Day 3 : 8:31:02 + revised SGD, Batch GD , Backpropagation Nov 8 Day 4: 14:00:00 (have to do leetcode and sql more tomorrow) Nov 9 Day 5: 17:13:31 (will learn CNN and CIFAR dataset project tomorrow) Nov 10 Day 6: 17:41:25 ( didn't feel like doing anything) Nov 11 Day 7: 1:01:36:57 (finished) Nov 12 not done yet! much to go! will finish the whole course now then move on to learning and understanding missing theory parts and building projects

[@godofeternals]

completed or not?

[@godofeternals]

should i start this course? is it worth it?

[@tsukuruuu]

​@@godofeternals Completed the course and wholeheartedly recommend it. The availability of book versions and comprehensive notebooks for revisions and code snippets has proven invaluable. It's one of the most well-structured and comprehensible courses out there. yes some portions are repetitive but that's also helpful for reinforcement or else you can just choose to skip over parts or, not watch video & read the book version and so on. whatever sails your boat. luck!

[@godofeternals]

@@tsukuruuu THANKS !!! will definitely do the course now (was pinned since 2 weeks) ill probably start after my OOPS exam tommorow T_T ;D

[@tsukuruuu]

@@godofeternals Good luck. i always disable comments, notifications , feed, shorts and every distracting thing on my browser so kinda replied late. Seems like you're in your second year. Good luck with examinations.

[@deepaksingh9318]

The more i watch of this the more i admire Daniel for his excellent teaching and presentation skills .. its exactly how the learner want the lesson to be "More Visuals and interactive".. Love how he explains each and every parameter in details , which i never see with other tutorials...

[@DanteNoguez]

I just finished the course and it's been absolutely wonderful. Thanks a lot for this, I look forward to see more of your content, Daniel!

[@mrdbourke]

Eyyy!! Massive effort Dante! Thank you for the kinds words & plenty more to come

[@angelmanuelvelascoreyes6741]

the best Pytorch course i have ever seen , thank you, i had to spend a long time trying to find a place to learn step by step and an interactive way as you have made your classes

[@mrdbourke]

Thank you Angel! Glad you’re enjoying it :)

[@sshubam]

Wow i remember when i watched your newly released tensorflow course on youtube in 2021, thankyou for making such resourceful content and I love your energy throughout the course! Good bless

[@mrdbourke]

If you liked the TensorFlow one, you’ll love this one!

[@yonrotem340]

Man I'm 8 hours in and loving this tutorial. Definitely helps having a great mentor and the familiar aussie accent for a change! Thanks for showing the entire process including hiccups. It is so often that you write code as a newbie and run into an error which is edited out in the video or tutorial you might follow. When this happens, panic ensues. So thanks! Only thing I would provide some feedback on is your song is fun and hilarious but i think to make it memorable it needs a hook. I'm happy to help you produce a beat for it (i'm a muso looking for a career in software development and AI)

[@timk2572]

Daniel is explaining it very well! So clearly. Thank you for that video!!! Looking for another ones. I am actually bought few courses on ZTM, one related to Tensor Flow

[@arnaudtremblay2243]

PyTorch lets you call zero_grad() manually because it is sometimes useful to store gradients before deleting them. An exemple would be a recurrent neural network (RNN).

[@linuxbrad]

11:00:00 entering the 11th hour of listening to this training at 75% speed during each of my outdoor runs. Some parts I've reviewed 4 times but it's all very digestible and articulated. You are amazing!

[@paolopiersanti806]

This tutorial is really great! When I cover the Deep Learning part in the course I will be teaching this Spring, I will definitely refer my students to your beautiful tutorial.

[@lcoandrade]

Great course!!!! The way you teach is really good. The best way is the "learn by doing" and you really do it!!!! Congrats!

[@danmold4483]

Thank you Sir for your amazing job! I've tried some free and paid courses on different study platforms and none of them isn't nearly close to your approach of teaching. As a rookie in data science and machine learning this is helping me a lot to get straight into the practice smoothly. My deep appreciation to you, Daniel

[@mrdbourke]

Thank you so much! Glad you’re enjoying it :)

[@user-oj4yl4tf6b]

Thanks, guy. It's great for you to provide this free course for pytorch. I had some experience on Tensorflow and scikit-learn and just want to go further on pytorch.

[@Kermin00]

Just finished it. A massive piece of great work. Thank you

[@amanjha9759]

Finally after a whole week I completed this and cannot thank you enough for the generosity. Thanks!!

[@mrdbourke]

Woohooo!!! Massive effort Aman!

[@fundoo203]

How much effort did it take for you complete this in a week? How long each day? I'm planning myself, pls help

[@amanjha5916]

@@fundoo203 sure but I'd like to let you know that I had some understanding about machine learning from Andrew NG's course and being an engineering sophomore math wasn't a problem. My daily watch time ranged from 2-4 hours (I tend to speed up my videos by 2x), rest I tried to spend at least 1 hr everyday with official pytorch tutorial available with it's docs. After dealing with my other commitments and assignments If I had any spare time left I'll watch Stanford's CNN (will recommend) lectures which made the section on CNN very easy to digest. So all in all before spending hours and hours here I tried to prepare myself for the content so that there wasn't much time wasted in repetition. Though I did struggle with the custom dataset section due to my lack of programming skills but I tried to force myself to understand or spend at least an hour before I called it quits. It's been about 10 days since I completed this and I haven't stopped exploring so a heads up from me, ML is very exciting you're in for a treat. All the best!!

[@kiana-computer]

🔥@@fundoo203 by spending 4 hours a day you can see the course and practice coding , wish you luck kzbin.infogaming/emoji/7ff574f2/emoji_u1f525.png

[@MalcolmJones-bossjones]

Just finished the course finally, and I wanted to say you are a REMARKABLE teacher, seriously!!! Thank you so much for this, well worth sticking through! I plan on looking at the other courses you have as well that go beyond this youtube video!!!

[@123greenland]

How many days/weeks you took to complete this 25 hrs course?

[@MalcolmJones-bossjones]

@@123greenland great question, I definitely needed to break it up due to work, family and other things, plus I tried to do a lot of the extra curricular things to get a better understanding of the material as well. I don’t remember the number exactly because it took it over a spread out amount of time, but let’s call it as least a month spending an hour-3 for each session. The beginning took the longest for me to just grasp the new concepts but once you hit image classification it all just started clicking thanks to the prerequisite material finally coming around full circle

[@123greenland]

Thank you 👍

[@nikhilmudgal8541]

Few months back I saw a pytorch tutorial cooking up in your repo. Can’t say how excited I am now it’s available. Thanks Daniel.

[@mrdbourke]

Thank you for following along Nikhil! Enjoy it!

[@anthonymiller6234]

This looks amazing. Awesome job. I've been looking for a resource like this for a while. I'll be tackling this in a couple of weeks. thanks

[@mrdbourke]

Thank you Anthony! Enjoy it when you do :)

[@ignaciobosch7006]

Daniel, thank you so much for this amazing contribution! Your own "transfer learning" about Pytorch (also TensorFlow) is amazing! Massive effort ;)

[@mrdbourke]

Thank you legend! Hahaha you’re right, turns out transfer learning works for people too!

[@ignaciobosch7006]

@@mrdbourke legend Mcgiver! haha doing the PyTorch course I cant avoid to thinking on you learning all this stuffs from scratch to teach us!

[@edeltobi]

I am now in the last hour of this course, and I want to thank you so much for this, Daniel. This is an incredible piece of education. I have been taking data science and programming courses for the last 6 years and quite frankly never had a boost in understanding, skills & context knowledge like this. You have really done many of us aspiring ML engineers and AI-savvy generalists a massive service by putting this together. I take my hat off and salute you for the time and effort. (but maybe dial down on the inofficial PyTorch song :D )

[@ewplayer3]

I’m into the last 1/4 of this course and I couldn’t agree more. There’s so much information out there on how to retrain prebuilt NNs, but very little on how to build one from the ground up (and what little exists is either limited or vague). I feel like I am taking a college level ML course while watching this. I’m learning more than I could have ever dreamed possible.

[@shafir360]

How long did you take you to complete the video?

[@ewplayer3]

@@shafir360 it took me about a month. It worked on it over my work lunches.

[@pekopeko3023]

about 2 weeks@@shafir360

[@Elon66688]

Thank you! I will comment again once I finish this course!

[@rafaared]

Thank you Daniel, great class :) almost finishing it.

[@hannanshaikh5939]

Thirteen Days completion results: 2:07:53 Day 1 (January 17) 3:57:31 Day 2 (January 18) 4:57:37 + exercises Day 3 (January 19) [was out for the whole day] 6:19:37 Day 4 (January 20) [took time to understand the codes cuz I am not a computer science student] 9:00:00 Day 5 (January 21) 10:35:27 Day 6 (January 22) :( 12:30:00 Day 7 (January 23) :/ 14:48:28 Day 8 (January 24) :\ 16:38:00 Day 9 (January 25) :| 18:26:02 Day 10 (January 26) :| 21:22:34 Day 11 (January 27) :] 24:10:24 Day 12 (January 28) :) 1:01:36:57 Day 13 (January 29) :) Completed Finally!!! A lot still remains......

[@mrdbourke]

Keep going! Big effort on the first day!

[@hannanshaikh5939]

​@@mrdbourke Thanks Daniel. Haven't found anything hard to digest until now. Appreciate your efforts man.

[@Augustus1003]

Good job dude, I completely skipped my first night because I was so excited for this course. Currently at CNN model part. Hope to finish this and cover the theory afterwards.

[@hannanshaikh5939]

​@@Augustus1003Thanks buddy. Awesome to hear you're diving in as well! Let's crush it together!

[@mlenthusiast6464]

Awesome. I was hoping for a video on pytorch from you. And here it is.... I haven't completed watching the video but am ready to learn something new.... thanks Daniel

[@mrdbourke]

I think you’ll find PyTorch very fun!

[@manuandrade6]

Haven't watched the course yet, but I know you've been working on this for a while. Cibgratulations and thanks a lot for the course!!!

[@sandeepm3545]

Thanks Daniel, I was a complete newbie in this task. But this teaching method of yours gave me very good confidence. I am not an expert but can code and model fairly well. There is no doubt that you will make a better next generation because of the knowledge you share. There is no doubt that you are a very efficient teacher. May God help you in this great work of yours. Thanks again.

[@watchingvideos9871]

this is amazing. My favorite feature about this course is how he subtly keeps reenforcing the fact that you need to be willing to look things up yourself to get good at coding.

[@sssaturn]

for my learning reference: 3:34:15 PyTorch tensors & NumPy 4:16:58 PyTorch WorkFlow 14:36:42 Introduction to obtaining dataset, setting up Pytorch Dataset, and loading dataset into DataLoader 15:08:24 Creating DataLoaders

[@yvankoabiloa9490]

1:12:59

[@anatoliikashchuk7527]

Great course! It is very interesting to see how you are trying to figure out the manual random seed. It gives a sense of live coding. I haven't finished the course yet, so I don't know if you explain it in detail later, but I can explain the random seed thing. One can think of a random number generator as a list of random numbers. When you call randn an element [0] from this list is returned. On the next call - element [1] and so on. When you specify manual seed the index in a "random list" resets and on every call, you get again elements [0], [1], [2] and so on. That's why you've got the same elements over and over again. Each consecutive call to randn after manual seed will return the same random numbers.

[@QuarktaschemitSenf]

WOW! Thank you so much Daniel! I switched job, now im programming with Python since 5 Months and get into a ComputerVision Library which is based on Pytorch. But i wanted to understand whats happening into the background.Again: Ty for your effort and making this available for free!

[@paulntalo1425]

This is a crazy resource. I have learnt n enjoyed alot from your previous courses esp the Tensorflow free and paid for course in KZbin and Udemy. I was able to complete my master's thesis with ease. Am excited about this new course series for PyTorch 💐👏🎉

[@mrdbourke]

So stoked to hear Paul! If you liked the TensorFlow course, you’ll love this one!

[@kallolchatterjee4373]

I have learned Deep Learning with Tensorflow course in Udemy from you. Really excited about this video. Thank you for uploading..

[@mrdbourke]

Enjoy Kallol! If you like that style, this one is much the same, except with PyTorch!

[@akashchandra2223]

@@mrdbourke when is the pytorch coming on udemy?

[@panceroti]

@@akashchandra2223 I would like to know that as well!

[@akashchandra2223]

@@panceroti already did

[@Das-wr5ou]

I mean, I just can't imagine how much work you have put in Daniel! Thank you very much. I have just finished watching. You are the best person!❤💯

[@aamirkhanmaarofi9705]

Wawo, its been six months I got back to this channel and it grew so much. Love the content you produce Daniel, more knowledge and power to you.

[@mrdbourke]

Thank you Aamir!

[@manassengudia1854]

I learned a lot of ML and DL concepts in an simplest way from Daniel ✊🏽 … I would like to see more content on transfer learning 🎉

[@mrdbourke]

Thank you Manasse! The full version of this course has PyTorch Transfer Learning as module 5: learnpytorch.io

[@yasthirdhewnarian6531]

I actually have a question Daniel. I bought your Tensorflow course last year and I just bought your PyTorch course now. Since you have done courses on both TF and PT, what is your opinion on the "which one is better" debate?

[@sunday7425]

Thanks for such an indepth PyTorch course, enjoyed it thoroughly

[@VishalSingh-qn7pp]

Thank you so much Daniel. This has helped me a lot. Never expect this good a course to be free. thanks a lot

[@ayky9389]

I finished the whole 25 hours. Thank you for your guide, Daniel! It really really helped me get started and wrap my head around PyTorch! I really enjoyed your teaching style, even though sometimes, especially in the last third, I had to pause a little often. Nevertheless, it was really great. I'd love to see going from classification to object detection now, with drawing bounding boxes and localizing multiple different objects in our image. Do you have something like this planned or already published?

[@sachinb_it1308]

congrats vro keep up the good work all thebest

[@msakdreams6374]

Thanks a lot sir. 1 week ago, I knew only 5% about machine learning and deep learning. After completing your 14+ hours TensorFlow course and 25+ hours PyTorch course, I have really learned a lot and got the proper guidelines how to keep everything together. To any newbie reading this, you just need a pc and patience to complete this course fluently. Thank you.

[@athenasu2736]

Only 30 mins in and hope I'm able to finish the entire video, and just wanted to comment that I like to think of embeddings as something that the computer needs to learn what humans think is "meaning" or "context." I thought that helped me a lot once I understood what the higher level of "finding patterns" meant. Happy learning everyone! And thanks for sharing Daniel!

[@cataori_program7292]

Thank you Daniel! Very enjoyable and earnest approach to the topic! personally, very useful.

[@raykeating3389]

Incredible teaching style and well thought out content. Thank you for posting this for free.

[@mrdbourke]

Thank you Ray! Glad you’re enjoying :)

[@Gap_cast]

It's weird but in chapter 41, I got [0.2345] too and it's even more weirder because when I changed my values I got the same final value again too OrderedDict([('weights', tensor([0.3367])), ('bias', tensor([0.1288]))]) as you got, and it's makes me wonder about randomness in PyTorch!! by the way, Thanks for your fantastic course it's brilliant.

[@karangirigauswami]

Same, After removing the line 'torch.manual_seed(42)', I consistently get the value 0.2345 on the first attempt.

[@jaysteele3731]

This is one of the best courses I’ve ever watched, can’t believe this is free. You are unreal Daniel

[@cyprianfusi8526]

Thank you very much Daniel, for your sacrifice and for sharing your knowledge with the world! GREAT content!!!

[@mbottambotta]

I'm two hours into the course. First, thank you Daniel for being so generous. Your course is a gold mine. Second, note to self: finish this! Everyone, thanks for holding me accountable 🤣

[@daniJohn847]

Have you finished it yet?

[@mbottambotta]

@@daniJohn847 thanks for asking! not finished, but working through the material. love it!

[@mbottambotta]

​@@daniJohn847 eight and a half hours in, still not given up. it's a challenge, for sure. but i'm holding on.

[@Michallote]

Currently 7 hours and a half in. Have you finished it jet? Hahaha

[@jamesbriggs]

Yes 🔥 looking forward to going through this - I'm curious, how long did it take you to pull all of this together?

[@mrdbourke]

Yo James! I worked out it takes about 1 week per hour of content (plus or minus a couple of days). The first commit to the repo was October last year. So this video itself probably took ~6 months a few hours per day. You can see the progress log here: github.com/mrdbourke/pytorch-deep-learning#log

[@hemanth8195]

Yoo broo

[@mrdbourke]

@@hemanth8195 yo brother long time no see!

[@jamesbriggs]

@@mrdbourke 1 hr == ~1 week sounds familiar - I like the log idea it's really cool, thanks for the commitment to this it's awesome and well worth the time with these results 👏🏼

[@hemanth8195]

@@mrdbourke ❤️❤️😍 l looking forward to see your more tut like this

[@jingrunchen7290]

Finished the video.🎉 Appreciate your hard work. Thx a lot

[@lukmanaj]

I'm thrilled to share my experience with the course. It's incredible how accessible and comprehensive this learning resource is, even for someone like me in Nigeria. The course has truly opened my eyes to the vast potential of open-source knowledge and PyTorch. I'm eager to see how this will shape the future of learning and innovation worldwide. Thank you, mrdbourke, for this amazing opportunity!

[@pranawmishra7330]

Hello Daniel, I am a big fan of your work and I really appreciate the effort you have put into this course. I just wanted to ask will this complete course be available on udemy and if yes can you tell us when??

[@mrdbourke]

Hey Pranaw! And yes it will be on Udemy soon (no official dates yet, when I’ve finished all the materials, I’ll upload them)

[@mrdbourke]

Hey hey Pranaw! Good news, it just launched today on Udemy: www.udemy.com/course/pytorch-for-deep-learning/?couponCode=ZTMGOODIES7 use code "ZTMGOODIES7" for a special launch discount (valid for the next ~4 days).

[@CreativeAiArtistry]

As am starting this deep learning course, I pray I complete this in Jesus name. wish me luck guys! and thank you Mr. Daniel Bourke for your wonderful intros and teaching style. I feel so torched to Pytorch🔥 and know so much more in the Deep Learning career.

[@amongus-rq2hb]

Completed the course. Thankyou Daniel for this awesome course

[@edos3783]

Hey Daniel thanks for the awesome tutorial, i'd like to ask a little question , at 3:40:40 you mentioned numpy array and pytorch tensors do not share memory although when I tried it myself for some reason the change of array didn't affect the tensor BUT the change of the tensor did affect the array, this is what i wrote and the output: #tensor to numpy array tensor = torch.ones(7) numpy_tensor = tensor.numpy() tensor ,numpy_tensor (tensor([1., 1., 1., 1., 1., 1., 1.]), array([1., 1., 1., 1., 1., 1., 1.], dtype=float32)) #tensor manipulation does change the numoy array! tensor +=1 tensor , numpy_tensor (tensor([2., 2., 2., 2., 2., 2., 2.]), array([2., 2., 2., 2., 2., 2., 2.], dtype=float32)) arr = np.arange(1.0 , 8.0) tensor = torch.from_numpy(arr) # numpy's default type is float64 , torch works with float32 so we need to convert when migrating data using numpy arr , tensor # array manipulation does not change the tensor ! arr += 1 (array([1., 2., 3., 4., 5., 6., 7.]), tensor([1., 2., 3., 4., 5., 6., 7.], dtype=torch.float64)) thanks for anyone who helps and have a great day !

[@jianfengjian6844]

The torch.Tensor.numpy() method returns a NumPy array that is a view of the underlying data of the tensor. It does not create a copy of the data but provides a new object that references the same memory.

[@tazeemmahashin7545]

I didn't even know it was possible to make a 25 our long video 😂 I'll eat this video up though, thank you Daniel!

[@mrdbourke]

Neither did I until I had some help from the kind people at freeCodeCamp. This took a few upload attempts but eventually it worked

[@_MrCode]

This course takes me from 0 ML/DL knowledge to a professional, I landed a job finally, Really appereciate his work && effort, He is a true gem among the AI instructors, he knows well how to teach almost anyone who wants to learn AI.

[@saulo_foot]

Now it seems I'll definitely learn pytorch and deep learning in a confident way. Thanks and congratulations!

[@hakankosebas2085]

beginner friendly, 1 day long video😅

[@aniruddhajana9987]

It is the best PyTorch Tutorial, I have ever seen. Thanks for giving so much content and knowledge for free. Hoping to see more tutorials like this in future. ❤❤

[@user-jm9lm6ot4e]

man you finally come back!!! can' t wait to see these amazing videos!!!

[@mrdbourke]

I’m back! This video just took a little while to make

[@amruthark9071]

Thank you for tutorial, its been really helpful!!. Cheers for doing a great work!!

[@alexandergeorgiev2631]

The reason why indentation matters at 5:34:40 is because in the first way he defined the forward function inside of the constructor, but you need to define the forward function as an instance method of the class so it's accessible.

[@zy5492]

Amazing course, amazing teacher! Thank you Daniel!!👏👏👏

[@Larimuss]

Thanks for putting this up! Really well explained.

[@avibcci1297]

Omg. Love you so much !!! Love you so much. I needed this so bad 😭

[@ZaidLyfe]

Dude!!!! such long form proper detailed tutorials are so helpful. OMG m speechless :)

[@mrdbourke]

Enjoy Zaid!