Menu
I am trying to explore model tuning through tensorboard profiling tab and was trying to generate files through tensorboard call back as shared below.
log_dir="logs/profile/" + datetime.now().strftime("%Y%m%d-%H%M%S")
tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=log_dir,
histogram_freq=1, profile_batch = 3)
model.fit(train_data,
steps_per_epoch=20,
epochs=10,
callbacks=[tensorboard_callback])
and it has generated following files in my colab. Have then downloaded these files into my local PC to view in tensorboard but nothing is getting displayed in Profile tab. All the other tab showing information.
logs/profile/ logs/profile/20190907-130136/ logs/profile/20190907-130136/train/ logs/profile/20190907-130136/train/events.out.tfevents.1567861315.340ae5d21d3b.profile-empty logs/profile/20190907-130136/train/events.out.tfevents.1567861301.340ae5d21d3b.119.129998.v2 logs/profile/20190907-130136/train/plugins/ logs/profile/20190907-130136/train/plugins/profile/ logs/profile/20190907-130136/train/plugins/profile/2019-09-07_13-01-55/ logs/profile/20190907-130136/train/plugins/profile/2019-09-07_13-01-55/local.trace
The script is located in https://colab.research.google.com/github/tensorflow/tensorboard/blob/master/docs/r2/tensorboard_profiling_keras.ipynb
Wanted to attach the files but there is no option to attach he files here...Can anyone please help why profile info from this script is not getting displayed into local PC tensorboard profile tab?
385
Load TensorBoard using Colab magic and launch it. View the performance profiles by navigating to the Profile tab. The performance profile for this model is similar to the image below. The Profile tab opens the Overview page which shows you a high-level summary of your model performance.
Use the TensorFlow Profiler to profile the execution of your TensorFlow code. The TensorFlow Profiler requires the latest versions of TensorFlow and TensorBoard ( >=2.2 ). Confirm that TensorFlow can access the GPU.
Machine learning algorithms are typically computationally expensive. It is thus vital to quantify the performance of your machine learning application to ensure that you are running the most optimized version of your model. Use the TensorFlow Profiler to profile the execution of your TensorFlow code.
Reusing TensorBoard on port 6006 (pid 750), started 0:00:12 ago. (Use '!kill 750' to kill it.) <IPython.core.display.Javascript object>
Apparently you need to use Chrome to view the profile information: https://github.com/tensorflow/tensorboard/issues/2874
138
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With
