Hi,
I just wanted to get some insights on whether TensorFlow is still the preferred framework these days and how updated the information of the course material is compared to now. Week 3 lab mentions it uses TensorFlow 2.3, which was released in 2020, therefore I am curious whether some views have changed since then. For example, the advantages of TensorFlow over e.g Pytorch. Mostly curious as to whether there is knew information that can be useful to add into the course’s materials and to also expand on the choice of framework discussion. Any insights are appreciated.
Have you seen this? Depending on your area of evaluating frameworks, consider exploring google trends as well.
Despite the improvements in tensorflow over time, the API part has been fairly stable till 2.14. TF 2.15 changes in APIs slightly since it started using keras3.
From the specialization viewpoint, the version of tensorflow in the labs is just fine as we code math expressions or use layers both of which have very few changes till date.
Thanks Balaji.
I see the it has many useful links to some online discussions. This is great! Is it possible to set this thread as duplicate and link to those discussions instead?
I think overall, (in addition to the frameworks - which is an important topic that week of the course), it just made me think whether there are some more readings outside the course or insights that have changed since the making of the videos - mostly because the world of deep learning / AI, etc. is advancing so fast. Mostly a reflection on the course in general. But i guess a lot of this can be found on the Batch newsletters if we go through them.
Thanks again for the reply.