All Questions
632 questions
0
votes
1
answer
40
views
What's necessary for the `keras` tag to populate on TensorBoard?
I'm trying to view the conceptual graph of a fairly complex TensorFlow model in TensorBoard. However, the option is greyed out. (I have no issue viewing the op graph).
My understanding is that in ...
0
votes
0
answers
60
views
How to set a cmap = 'hsv' in tensorboard logger add_image?
I have one channel images and want to add them to the tensorboard logger by using writer.add_image, as follows:
generated_images_to_show = torch.cat(gtimage_cpu, predimage_cpu), dim=0)
...
0
votes
1
answer
129
views
How can I download charts as SVG from TensorBoard?
I want to download the charts generated by TensorBoard as a vector graphics file so that I can include these as plots in documents and PowerPoint presentations. I have a vague memory that this has ...
0
votes
2
answers
372
views
PyTorch TensorBoard SummaryWriter giving empty metrics
I've been setting up tensorboard with PyTorch. Originally I was using lightning, and ran into an issue where I would log my hyperparameter, but the metrics in the HPARAMS tab would be empty. So, to ...
0
votes
1
answer
113
views
add_hparams() function from TensorBoard doesn't work properly
I'm trying to add metrics to this SummaryWriter, but it's not working.
I'm using the add_hparams() function of SummaryWriter, the details of which can be found here: https://pytorch.org/docs/stable/...
0
votes
0
answers
45
views
Can't get full data log from TensorBoard (Only 1000)
When I download the scalar data from TensorBoard, it only exports 1000 steps.
The tutorial online requires modifying "scalar_metadata.PLUGIN_NAME" in application.py.
However, after I opened ...
1
vote
1
answer
100
views
The conflict is caused by: The user requested tensorboard==1.12.2 and tensorflow==1.12.0
PS D:\New folder (2)\TamilChatbot-banking-service\Chatbot-banking-service-master\rasa_nlu> pip install -r requirements.txt
Defaulting to user installation because normal site-packages is not ...
1
vote
1
answer
647
views
Tensorboard profile : No profile data was found
Problem
I dont see anything in the profile section in Tensorboard. I get the following Tensorboard interface after running tensorboard --logdir logdir
The tree of logdir is as follow:
logdir
├── ...
-1
votes
2
answers
1k
views
How do I access to the logs generated from training a model with tensorboard?
now That we cannot upload the logs to tensorboard dev I cannot transform the information from the logs to DataFrame as easily as before. So I wante to ask if anyone knows how to extract the data ...
1
vote
0
answers
500
views
Need Help Reading the TensorBoard Profiler Tea Leaves : TensorFlow data.Dataset.prefetch with AUTOTUNE Running Synchronously
I'm trying to figure out how to improve the data pipeline performance for a CNN model that I am running on Google Cloud with the dataset stored in Google Cloud Storage. The image dataset is too large ...
1
vote
0
answers
114
views
TensorBoard graph is ugly
I hope someone had the same problem. My TensorBoard graph looks very cluttered, and I'm having trouble understanding it. I'm trying to implement a neural network for denoising, but I'm wondering if ...
0
votes
1
answer
144
views
Tensorboard histogram onehot operation causing ResourceExhauseError: OOM
I'm trying to train a VGG16 model. I'm using a sample dataset of 4000 300x300 images in 14 classes, and running my code on a Google VM using an Nvidia L4 GPU with 20gb of memory. I am running python 3....
3
votes
0
answers
624
views
Tensorboard profiler tab not showing anything behind corporate firewall
Versions: python3.7, tensorflow 2.9.1, tensorboard 2.9.1, tensorboard-plugin-profile==2.11.2.
tb_callback is passed in as follows
tb_callback = tf.keras.callbacks.TensorBoard(logdir=path_tologdir, ...
0
votes
1
answer
2k
views
Using keras tuner for hyperparameter tuning and extracting trial statistics with Tensorboard workaround
I'm using keras tuner for hyperparameter tuning my sequential neural network in keras. My aim is to let keras tuner do a lot of trials and then save all statistics - loss, hyperparameters, trial ...
4
votes
0
answers
536
views
How to visualize timeseries output at each epoch in tensorboard
I have a keras model that outputs an MxN tensor along with several scalar metrics. The metrics show up just fine in the SCALARS and TIME SERIES tabs:
This view is great for visualising scalars where ...