Develop Reference

Is there a simple way to know which Tensorflow ops have a registered GPU kernel?

I have been trying to optimize some Tensorflow code that was pretty memory inefficient (use of large dense tensors containing very sparse information), and would thus limit batch size and scalability, by trying to make use of SparseTensors.
After some struggle I finally come up with a decent solution with satisfactory speedup on CPU and very low memory usage, and when the time comes to use a GPU I realize that the previous memory inefficient is orders of magnitude faster...
Using tensorboard profiling I've discovered that two of the operations I have used in my ""optimized"" version only run on CPU (namely UniqueV2 and sparse_dense_matmul), but I could not see any hint of that in the documentation.
The only related piece of documentation states:
If a TensorFlow operation has no corresponding GPU implementation,
then the operation falls back to the CPU device. For example, since
tf.cast only has a CPU kernel, on a system with devices CPU:0 and
GPU:0, the CPU:0 device is selected to run tf.cast, even if requested
to run on the GPU:0 device.
In turn there is nothing in the tf.cast documentation hinting that the op has no GPU kernel.
Thus, is there a simple way to know whether a TF ops has a registered GPU kernel, without having to use a GPU to find it out?
The custom ops guide suggest that this could be seen by looking at the ops C files, but this seems a rather cumbersome way to do it...
I'm using TF v2.8
Thanks!

Erratic GPU Utilization while training GAN in Tensorflow 2.5

I'm using Tensorflow 2.5 to train a starGAN network for generating images (128x128 jpeg). I am using tf.keras.preprocessing.image_dataset_from_directory to load the images from the subfolders.
Additionally I am using arguments to maximize loading performance as suggested in various posts and threads such as loadedDataset.cache().repeat.prefetch
I'm also using the num_parallel_calls=tf.data.AUTOTUNE for the mapping functions for post-processing the images after loading.
While training the network on GPU the performance I am getting for GPU Utilization is in the picture attached below.
My question regarding this are:
Is the GPU utlization normal or is it not supposed to be so erratic for traning GANs?
Is there any way to make this performance more consistent?
Is there any way to improve the training performance to fully utlize the GPU?
Note that Ive logged my disk I/O also and there is no bottleneck reading/writing from the disk (nvme ssd).
The system has 32GB RAM and a RTX3070 with 8GB Vram. I have tried the running it on colab also; but the performance was similarly erratic.

It is fairly normal for utilization to be erratic like for any kind of parallelized software, including training GANs. Of course, it would be better if you could fully utilize your GPU, but writing software that does this is challenging and becomes virtually impossible when you are talking about complex applications like GANs.
Let me try to demonstrate with a trivial example. Say you have two threads, threadA and threadB. threadA is running the following python code:
x = some_time_comsuming_task()
y = get_y_from_threadB()
print(x+y)
Here threadA is performing lots of calculations to get the value for x, retrieving the value for y, and printing out the sum of x+y. Imagine threadB is also doing some kind of time consuming calculation to generate the value for y. Unless threadA is ready to retrieve y at the exact same time threadB finishes calculating it, you won't have 100% utilization of both threads for the entire duration of the program. And this is just two threads, when you have 100s of threads working together with multiple chained data dependencies, you can see how it becomes exponentially more difficult to eliminate any and all time threads spend waiting on other threads to deliver input to the next step of computation.
Trying to make your "performance more consistent" is pointless. Whether your GPU utilization went up and down (like in the graph you shared) or it stayed exactly at the average utilization for the entire execution would not change the overall execution time, which is probably the actually important metric here. Utilization is mostly useful to identify where you can optimize your code.
Fully utilize? Probably not. As explained in my answer to question one, it's going to be virtually impossible to orchestrate your GAN to completely remove bottlenecks. I would encourage you to try and improve execution time, rather than utilization, when optimizing your GAN. There's no magic setting that you're missing that will completely unlock all of your GPU's potential.

Is there a way to run tf.data API on GPU in tf >= 2.4

I am surprised that I could not see a concise way to run a tf.data API on the GPU. I understand that the data pipelines can run on CPU so that it can happen in parallel (with pre-fetching), allowing the GPU to run actual model and train it.
However, my pre-processing is extremely parallel and computationally intensive. While I can technically write the pre-processing as the first layer in my model, I would really not prefer to do this to prevent training data leakage into my model.
Any pointers is appreciated for this. The closest I found was https://towardsdatascience.com/overcoming-data-preprocessing-bottlenecks-with-tensorflow-data-service-nvidia-dali-and-other-d6321917f851, which involves using nvidia DALI frameworks.
Here are some critical points:
I have already tried enforcing device placement with tf.device('...').
I dont want to prefectch the data into the device but rather run the whole data pipeline on GPU.
Preferably, if my computation is more, I want to save my dataset as a tfrecords, so that I can load it directly. This can be done with tf.data.experimental.save for now, but it again it uses CPU!

Tensorflow not fully utilizing GPU in GPT-2 program

I am running the GPT-2 code of the large model(774M). It is used for the generation of text samples through interactive_conditional_samples.py , link: here
So I've given an input file containing prompts which are automatically selected to generate output. This output is also automatically copied into a file. In short, I'm not training it, I'm using the model to generate text.
Also, I'm using a single GPU.
The problem I'm facing in this is, The code is not utilizing the GPU fully.
By using nvidia-smi command, I was able to see the below image
https://imgur.com/CqANNdB

It depends on your application. It is not unusual to have low GPU utilization when the batch_size is small. Try increasing the batch_size for more GPU utilization.
In your case, you have set batch_size=1 in your program. Increase the batch_size to a larger number and verify the GPU utilization.
Let me explain using MNIST size networks. They are tiny and it's hard to achieve high GPU (or CPU) efficiency for them. You will get higher computational efficiency with larger batch size, meaning you can process more examples per second, but you will also get lower statistical efficiency, meaning you need to process more examples total to get to target accuracy. So it's a trade-off. For tiny character models, the statistical efficiency drops off very quickly after a batch_size=100, so it's probably not worth trying to grow the batch size for training. For inference, you should use the largest batch size you can.
Hope this answers your question. Happy Learning.

Modify Tensorflow Code to place preprocessing on CPU and training on GPU

I am reading this performance guide on the best practices for optimizing TensorFlow code for GPU. One suggestion they have is to place the preprocessing operations on the CPU so that the GPU is dedicated for training. To try to understand how one would actually implement this within an experiment (ie. learn_runner.run()). To further the discussion, I'd like to consider the best way to apply this strategy to the Custom Estimator Census Sample provided here.
The article suggests placing with tf.device('/cpu:0') around the preprocessing operations. However, when I look at the custom estimator the 'preprocessing' appears to be done in multiple steps:
Line 152/153 inputs = tf.feature_column.input_layer(features, transformed_columns) & label_values = tf.constant(LABELS) -- if I wrapped with tf.device('/cpu:0') around these two lines would that be sufficient to cover the 'preprocessing' in this example?
Line 282/294 - There is also a generate_input_fn and parse_csv function that are used to set up input data queues. Would it be necessary to place with tf.device('/cpu:0') within these functions as well or would that basically be forced by having the inputs & label_values already wrapped?
Main Question: Which of the above implementation suggestions is sufficient to properly place all preprocessing on the CPU?
Some additional questions that aren't addressed in the post:
What if the machine has multiple cores? Would 'cpu:0' be limiting?
The post implies to me that by wrapping the preprocessing on the cpu, the GPU would be automatically used for the rest. Is that actually the case?
Distributed ML Engine Experiment
As a follow up, I would like to understand how this can be further adapted in a distributed ML engine experiment - would any of the recommendations above need to change if there were say 2 worker GPUs, 1 master CPU and a parameter server? My understanding is that the distributed training would be data-parallel asynchronous training so that each worker will be independently iterating through the data (and passing gradients asynchronously back to the PS) which suggests to me that no further modifications from the single GPU above would be needed if you train in this way. However, this seems a bit to easy to be true.

MAIN QUESTION:
The 2 codes your placed actually are 2 different parts of the training, Line 282/294 in my options is so called "pre-processing" part, for it's parse raw input data into Tensors, this operations not suitable for GPU accelerating, so it will be sufficient if allocated on CPU.
Line 152/152 is part of the training model for it's processing the raw feature into different type of features.
'cpu:0' means the operations of this section will be allocated on CPU, but not bind to specified core. The operations allocated on CPU will run in multi-threads and use multi-cores.
If your running machine has GPUs, the TensorFlow will prefer allocating the operations on GPUs if the device is not specified.

The previous answer accurately describes device placement. Allow me to provide an answer to the questions about distributed TF.
The first thing to note is that, whenever possible, prefer a single machine with lots of GPUs to multiple machines with single GPUs. The bandwidth to parameters in RAM on the same machine (or even better, on the GPUs themselves) is orders of magnitude faster than going over the network.
That said, there are times where you'll want distributed training, including remote parameter servers. In that case, you would not necessarily need to change anything in your code from the single machine setup.