DataBloom

Hi, I am looking at Tensorflow tutorials and would like some opinions on the best free Tensorflow course. I have looked at these tutorials so far:

freeCodeCamp Tensorflow complete tutorial https://youtu.be/tPYj3fFJGjk

Udacity Tensorflow tutorial https://www.udacity.com/course/intro-to-tensorflow-for-deep-learning–ud187

However, these are 1-2 years old. Do they still hold up and which of the two should I pick?

Feel free to suggest other tutorials too!

Hey all!

I’m trying to implement this Transformer article. In the source code they pass a number (n_heads) representing the number of attention heads that should be created, which is used when building the model to create that number of attention heads and save them to a list. Later, when the model is called, the attention heads are iterated over as follows: attn = [self.attn_heads[i](inputs) for i in range(self.n_heads)]. When running this code in graph mode the following error is thrown: OperatorNotAllowedInGraphError: iterating over “tf.Tensor” is not allowed: AutoGraph did convert this function.

How should I go about creating an arbitrary number of the same layer in such a way that it can be iterated over in graph mode?

Is there an age requirement for the tensorflow developer certificate exam?

TensorFlow was working fine when I only had a 1660ti, but I recently installed a 3070 and the 3070 works fine for other stuff, but not TensorFlow.

It will take a good minute to just do – tf.test.gpu_device_name()

import os os.environ["CUDA_VISIBLE_DEVICES"]="0" import tensorflow as tf tf.test.gpu_device_name() 

If CUDA_VISIBLE_DEVICES is changed to “1” (my old card) it will print the device name instantly.

Everything else TensorFlow related will also take super long on the new card.

Is there some box I have to check in the CUDA installation files for the new card to work, or do I need to specially install something to find the new card?

I’m using Jupyter Notebook.

Any help would be appreciated.

I was working on an independent project and our school only has AMD GPU with Mac. (we have NVIDIA GPU PC but it is only GTX1030)

Hey everyone 👋,

I have been taking part solo in an ML challenge by AIcrowd.com which has a cash prize pool of $50,000 🤑

The Machine Learning challenge is for Object Detection enthusiasts, hosted by Amazon Air Prime, called Airborne Object Tracking

The challenge revolves around predicting the future motion of flying airborne objects to avoid collision. Also the dataset they have is pretty lit (11TB in size!! 😲), one of the largest collections of flight sequences from aerial vehicles so you might want to look into that.

I am looking for someone to team up in this challenge with. Let me know in the comments if anyone is up!

I have found the official introduction example guides on the official site https://www.tensorflow.org/tutorials

But I could not find more example codes up until I have bumped in to: https://www.tensorflow.org/text/tutorials/

So is there any other example full projects specialized on topics. (I love their format and the added COLAB just makes it easier to learn it.)