PyTorch is what I made a transition to from Tf2.x(x>=7). So far I see it quite cool & friendly to research community. And yes, I converted my entire TF code to PyTorch in less than 2weeks.
TF usually has some features 1-2 year ahead pytorch. Pytorch is more flexible for tweaking the model. TF2.4-2.6 was very buggy with strange errors that took long time to fix. That was the time I switched to mainly pytorch. I think TF is better now.
Your reply was a year ago but you seem to be very familiar with the 2 frameworks. I am starting out with one of the two. I am interested in vision based research and academia. Which one should I pick? Thanks in advance!
Both are great. TF with Keras has better performance, strong community support and robustness 👍PyTorch better for research and experimentation. Easier to use and debug since is more pythonic. Better Dynamic computation graph. Easier to deploy on web and mobile. Your choice will be determine on: use case and developer's preferences.
In other words, TF's source code is complete garbage with no forethought taken? Jeff Dean just sat down one day and started typing until a framework emerged. Got it...
@@cristian-bull I'm sure you can give examples of that, as well as examples of when TF's source code lesser elegance makes it less effective at run time or solving a bug or adding features?
@@noli-timere-crede-tantum When someone tells you they like pepperoni pizza, do you say "oh... so what you're saying is that ham and pineapple pizza is garbage?!". No one is saying what you are accusing them of saying. Here's a better, more productive pathway. Why don't you tell us why you think tensorflow is better, instead of asking everyone else to do that for you.
I like calling pytorch imperative! Because it you would understand machine learning in better way! Unlike declarative approaches similar to Unix like command and SQL you give it a command or query and it would run. Even through that Tensorflow uses procedural language, coding with it similar to declaring command. And static graph structuring making it difficult especially for beginners debugging code!
More like collaboration than competition. The open source _COMMUN_-ity shows how a _commune_ environment that _shares_ resources can be so vastly superior to the capitalistic wealth redistribution scheme, it isn't even funny.
Pretty useless clip tbh, all he says is "I prefer Pytorch because I've been using it for longer". He mentions that the imperative style is easier to debug, but TensorFlow 2 also uses an imperative style.
I think his initial disclaimer completely disqualifies him from answering the question. If he hasn't built something meaningful in TF2, how can he make any comparisons whatsoever?
This content is incredibly moving. I read a book with akin material that reshaped my worldview. "AWS Unleashed: Mastering Amazon Web Services for Software Engineers" by Harrison Quill
![เป็นไปได้ไหม - WanMai [Performance Video]](http://i.ytimg.com/vi/Eh5hItD4Mr0/mqdefault.jpg)
Asking a Facebook employee about pytorch vs tensorflow. 🤔
Yup ..
Lol
They make the largest no of open-source AI models so why not.
They have currently 1907models.Refer Hugging Face
😂
PyTorch is what I made a transition to from Tf2.x(x>=7). So far I see it quite cool & friendly to research community. And yes, I converted my entire TF code to PyTorch in less than 2weeks.
TF usually has some features 1-2 year ahead pytorch. Pytorch is more flexible for tweaking the model. TF2.4-2.6 was very buggy with strange errors that took long time to fix. That was the time I switched to mainly pytorch. I think TF is better now.
That pretty much sums it up
Your reply was a year ago but you seem to be very familiar with the 2 frameworks. I am starting out with one of the two. I am interested in vision based research and academia. Which one should I pick? Thanks in advance!
@@sidgirase Pytorch
Both are great. TF with Keras has better performance, strong community support and robustness 👍PyTorch better for research and experimentation. Easier to use and debug since is more pythonic. Better Dynamic computation graph. Easier to deploy on web and mobile. Your choice will be determine on: use case and developer's preferences.
I mean there are more applications with PyTorch for computer vision purposes over tensorflow
For me pytorch can give you more granular control with small learning curve, but when it comes to deplotment and documentation tf is way ahead.
PyTorch’s source code is elegant and well thought out.
In other words, TF's source code is complete garbage with no forethought taken? Jeff Dean just sat down one day and started typing until a framework emerged. Got it...
@@noli-timere-crede-tantum Pytorch's source code is more elegant and better thought out.
@@cristian-bull I'm sure you can give examples of that, as well as examples of when TF's source code lesser elegance makes it less effective at run time or solving a bug or adding features?
@@noli-timere-crede-tantum When someone tells you they like pepperoni pizza, do you say "oh... so what you're saying is that ham and pineapple pizza is garbage?!". No one is saying what you are accusing them of saying. Here's a better, more productive pathway. Why don't you tell us why you think tensorflow is better, instead of asking everyone else to do that for you.
@@generichuman_ would PyTorch better for startups?
RIP tensorflow
Pytorch Forever 🔥🔥🔥 ... I know the XLA TPU support isn't almost there yet but yeah Pytorch 🔥
Using TPU very niche thought
PyTorch has more applications over tensorflow
pytorch is more intuitive for SWEs, tensorflow is killer now with the keras integration
Tensorflow...!
Great support, model integration and model deployment.
I like calling pytorch imperative! Because it you would understand machine learning in better way! Unlike declarative approaches similar to Unix like command and SQL you give it a command or query and it would run. Even through that Tensorflow uses procedural language, coding with it similar to declaring command. And static graph structuring making it difficult especially for beginners debugging code!
the key he is saying is which open source code is avail free for download
Thank you
More like collaboration than competition. The open source _COMMUN_-ity shows how a _commune_ environment that _shares_ resources can be so vastly superior to the capitalistic wealth redistribution scheme, it isn't even funny.
After keras integration tf is winner
wow 😀 ! you are diverse in your topics ! luv 'it 😘!
Now ask a Google employee the same question
Tensorflow is bread and butter
so it's boring?
Pretty useless clip tbh, all he says is "I prefer Pytorch because I've been using it for longer". He mentions that the imperative style is easier to debug, but TensorFlow 2 also uses an imperative style.
doesn't answer the question at all :|
I think his initial disclaimer completely disqualifies him from answering the question. If he hasn't built something meaningful in TF2, how can he make any comparisons whatsoever?
read between the lines
@@noli-timere-crede-tantum Have you used Pytorch?
This content is incredibly moving. I read a book with akin material that reshaped my worldview. "AWS Unleashed: Mastering Amazon Web Services for Software Engineers" by Harrison Quill
jax
WHY THE FUCK IS LEX FRIDMAN TALKING ABOUT MACHINE LEARNING
He is an AI PhD
#throwbacksundays
Interesting
Indeed
Rather
Very respectfully, but it's quite hard to understand what he is telling.
trrrrrrrrrrrrr