Closed vcoyette closed 3 years ago
Hi!
Thanks for taking your time to track this down :)
Your approach how to handle 4-dimensional image tensors is good, but Neptune server requires X-axis to be strictly monotonic (for each x, it must be larger than the previous one), otherwise all images except the first one will be dropped with a warning.
This particular piece of code should work with previews versions of TF as well - with 2 and 3 dimensional tensors
We created #41 based on your issue and sent it to review/testing on our side, you're welcome to take part in this process and see if it works with your code correctly :)
Hi,
We released version 0.5.1 that contains a fix based on your solution.
Feel free to try it out :)
Hi,
Sorry I missed you last comment. It is working great for me ! Many thanks for your efficiency :)
Hi Neptune team,
I am getting an error when logging an image to tensorboard after runnig neptune_tb.integrate_with_tensorflow().
The problem is that an image has to be of dimension 4 (k, h, w, c) when logged to tensorboard (see https://www.tensorflow.org/api_docs/python/tf/summary/image).
However, when tf.summary.image is called with an image of dimension 4 (shape=(1, 1200, 1200, 3) in my case), the following error is raised:
I think the limitation here is that neptune.experiment.Experiment.log_image only accepts one image. As a workaround, I replaced line 207 of neptune_tensorboard/integration/tensorflow_integration.py by:
I can create a PR if you want.
Versions: