Issue with opening Tensorboard with AWS sage maker

[AsifRazak-MohammedAbdul]

I am trying to open Tensorboard using AWS sage maker. There is an example code already available in AWS ( Link https://docs.aws.amazon.com/sagemaker/latest/dg/studio-tensorboard.html)

To Test I have followed the steps and opened tensorboard in aws host as above. I could see my Tensorboard but No data is displayed. The board is just empty (see the image)

I also wanted to use the PROJECTOR. So I created checkpoints during the training and saved it in the same path as per the Tensorflow documentation.

My log directory is test_logs/fit/20230412-215858 So, In the root directory of my Terminal - I used
tensorboard --logdir test_logs/fit/20230412-215858

But when i open the Tensorboard PROJECTOR i get an error (No Checkpoint was found)

I do not understand why I am receiving this error? The log_dir is correctly checked and I am in my root directory. Can anyone tell me what I am doing wrong?

[rileyajones]

I believe your logdir is incorrect. Can you try running tensorboard --logdir test_dir from the root dir?

These are the steps I followed to get this working:

1. Setup an environment mkdir 6313 && cd 6313 python -m venv 6313 && source 6313/bin/activate pip install tensorflow pip install tensorboard
2. Get the tutorial code I just copied all the code snippets from https://docs.aws.amazon.com/sagemaker/latest/dg/studio-tensorboard.html into a file which I titled 6313.py
3. Run the tutorial python 6313.py
4. Start tensorboard tensorboard --logdir logs

I'm not sure exactly what you did to try to get the projector working and without more details I can't help you.