-
Notifications
You must be signed in to change notification settings - Fork 32
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Insufficient documentation #7
Comments
Hey Antonio, sorry to hear you're running into issues. Are you running into any specific errors that could be documented? |
No, unfortunately no. My problem is that I do not know how to run this line: |
Ah, I see how those instructions could be confusing. In the instructions you linked, check out the code underneath this comment: "# Create options to profile the time and memory information." That will show you how to create a profile context that can be displayed in the UI. |
I already did it. If I execute my testing script from the terminal ( |
Now that you have a profile file, you can continue to Installation step 4: |
Installation step 3 has been rewritten based on what you shared in this bug, I hope it's clearer now, thanks |
Yes, it is clearer now. Thanks Chris. Following those instructions, however, I only get the report in the CLI. But how can I visualize the profiler on my web browser? |
Glad to hear that, Antonio. Now that this repository's (tensorflow/profiler-ui) instructions have been clarified, let's put the other repository's (tensorflow/tensorflow) instructions aside for now, as they focus on the CLI. The next step should be running this command from Step 4: |
No, it doesn't. Should it open automatically? When using TensorBoard, I have to open it manually from Chrome via localhost. I am using Chrome on a Mac, BTW. |
Yes, it should be opening automatically. Could I ask what's printing in your terminal after you run that command? |
Sure. There you go: 25 ops no flops stats due to incomplete shapes. =========================Options============================= ==================Model Analysis Report====================== Doc: Profile: ======================End of Report========================== =========================Options============================= ==================Model Analysis Report====================== Doc: Profile: ======================End of Report========================== =========================Options============================= ==================Model Analysis Report====================== Doc: Profile: ======================End of Report========================== =========================Options============================= ==================Model Analysis Report====================== Doc: Profile: ======================End of Report========================== Model accuracy is: 0.917 |
Interesting, could you send the command that generated this? |
|
So what if you ran this command instead? python ui.py --profile_context_path=/path/to/your/profile.context Where |
Finally! Now I get it. So this is procedure (please, correct me if I am wrong):
Thank you very much ChrisAntak. Nevertheless, I suggest that you add a brief model (e.g. MNIST for beginners) with the profile context already configured in it and a thorough step-by-step description of how to enter into the GUI of the profiler. It would help a lot. |
Glad to hear it's working for you! And thanks for those ideas. |
This TensorFlow Profiler UI looks promising but documentation is too brief. I installed
pprof
andbazel
. I managed to make it work in the CLI following these instructions. But I could not make it on the web browser. Please, give us some more details.The text was updated successfully, but these errors were encountered: