Is it just me or is tensorflow less popular than pytorch? Why is that?
", "messages": [ { "id": "1179132", "postDate": "01/31/2021 10:49:03", "content": "Is it just me or is tensorflow less popular than pytorch? Why is that?
", "rawMarkdown": "Is it just me or is tensorflow less popular than pytorch? Why is that?", "votes": null }, { "id": "1179279", "postDate": "01/31/2021 13:24:11", "content": "Tensorflow is backed by google and pytorch is backed by facebook. After Tensorflow 2.0 things are looking good again for tensorflow. You can try both to see which you like and want to pursue further.
", "rawMarkdown": "Tensorflow is backed by google and pytorch is backed by facebook. After Tensorflow 2.0 things are looking good again for tensorflow. You can try both to see which you like and want to pursue further.", "votes": null }, { "id": "1182898", "postDate": "02/02/2021 16:30:03", "content": "Pytorch has very clear syntax compared to tensorflow(model fitting is simpler in tf), tensorflow added/removed features from tf 1.0 to 2.0, so it is still a bit messy. On the other hand pytorch implementation of automatic differentiation much cleaner and easy to use, For beginners keras/tf is easy, but you should also get familiar with pytorch as many new implementations are in pytorch.
", "rawMarkdown": "Pytorch has very clear syntax compared to tensorflow(model fitting is simpler in tf), tensorflow added/removed features from tf 1.0 to 2.0, so it is still a bit messy. On the other hand pytorch implementation of automatic differentiation much cleaner and easy to use, For beginners keras/tf is easy, but you should also get familiar with pytorch as many new implementations are in pytorch.", "votes": null }, { "id": "1183960", "postDate": "02/03/2021 10:15:30", "content": "Pytorch is much more flexible as far as I feel. Everything will be in developer's hand. But on the other hand, tensorflow is a bit hard to make custom things in it. But still tensorflow has it's own advantages. I used to code in tensorflow but recently, I switched to pytorch. So, I felt pytorch is a bit comfortable!
", "rawMarkdown": "Pytorch is much more flexible as far as I feel. Everything will be in developer's hand. But on the other hand, tensorflow is a bit hard to make custom things in it. But still tensorflow has it's own advantages. I used to code in tensorflow but recently, I switched to pytorch. So, I felt pytorch is a bit comfortable!", "votes": null }, { "id": "1184324", "postDate": "02/03/2021 13:43:25", "content": "It is always better to learn more than one framework. I am currently using tensorflow. Can anyone tell as what would be the best resource in terms of book or online course to learn pytorch?
", "rawMarkdown": "It is always better to learn more than one framework. I am currently using tensorflow. Can anyone tell as what would be the best resource in terms of book or online course to learn pytorch?", "votes": null } ], "comments": [ { "id": 1179279, "author_name": "vickygoyal", "author_url": "", "post_date": "01/31/2021 13:24:11", "content": "Tensorflow is backed by google and pytorch is backed by facebook. After Tensorflow 2.0 things are looking good again for tensorflow. You can try both to see which you like and want to pursue further.
", "votes": null, "replies": [] }, { "id": 1182898, "author_name": "anku5hk", "author_url": "", "post_date": "02/02/2021 16:30:03", "content": "Pytorch has very clear syntax compared to tensorflow(model fitting is simpler in tf), tensorflow added/removed features from tf 1.0 to 2.0, so it is still a bit messy. On the other hand pytorch implementation of automatic differentiation much cleaner and easy to use, For beginners keras/tf is easy, but you should also get familiar with pytorch as many new implementations are in pytorch.
", "votes": null, "replies": [] }, { "id": 1183960, "author_name": "saimanojakondi", "author_url": "", "post_date": "02/03/2021 10:15:30", "content": "Pytorch is much more flexible as far as I feel. Everything will be in developer's hand. But on the other hand, tensorflow is a bit hard to make custom things in it. But still tensorflow has it's own advantages. I used to code in tensorflow but recently, I switched to pytorch. So, I felt pytorch is a bit comfortable!
", "votes": null, "replies": [] }, { "id": 1184324, "author_name": "vickygoyal", "author_url": "", "post_date": "02/03/2021 13:43:25", "content": "It is always better to learn more than one framework. I am currently using tensorflow. Can anyone tell as what would be the best resource in terms of book or online course to learn pytorch?
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1179132": "Is it just me or is tensorflow less popular than pytorch? Why is that?", "1179279": "Tensorflow is backed by google and pytorch is backed by facebook. After Tensorflow 2.0 things are looking good again for tensorflow. You can try both to see which you like and want to pursue further.", "1182898": "Pytorch has very clear syntax compared to tensorflow(model fitting is simpler in tf), tensorflow added/removed features from tf 1.0 to 2.0, so it is still a bit messy. On the other hand pytorch implementation of automatic differentiation much cleaner and easy to use, For beginners keras/tf is easy, but you should also get familiar with pytorch as many new implementations are in pytorch.", "1183960": "Pytorch is much more flexible as far as I feel. Everything will be in developer's hand. But on the other hand, tensorflow is a bit hard to make custom things in it. But still tensorflow has it's own advantages. I used to code in tensorflow but recently, I switched to pytorch. So, I felt pytorch is a bit comfortable!", "1184324": "It is always better to learn more than one framework. I am currently using tensorflow. Can anyone tell as what would be the best resource in terms of book or online course to learn pytorch?" }, "source": "meta" }