# tensorboard **URL:** https://discuss.pytorch.org/c/tensorboard/28.md [Latest](https://discuss.pytorch.org/latest.md) · [Categories](https://discuss.pytorch.org/categories.md) --- ## [About the tensorboard category](https://discuss.pytorch.org/t/about-the-tensorboard-category/96405)
**Author:** [@albanD](https://discuss.pytorch.org/u/albanD)\ **Replies:** 0
--- ## [Why is the data not converted to the CHW format of tensors after the ToTensor() function is used in torch 2.10, but instead remains in the HWC format？](https://discuss.pytorch.org/t/why-is-the-data-not-converted-to-the-chw-format-of-tensors-after-the-totensor-function-is-used-in-torch-2-10-but-instead-remains-in-the-hwc-format/224445)
**Author:** [@kjbusion823](https://discuss.pytorch.org/u/kjbusion823)\ **Replies:** 1\ **Last updated:** [January 30, 2026, 1:28pm UTC](https://discuss.pytorch.org/t/why-is-the-data-not-converted-to-the-chw-format-of-tensors-after-the-totensor-function-is-used-in-torch-2-10-but-instead-remains-in-the-hwc-format/224445 "2026-01-30T13:28:04Z")
Why is the data not converted to the CHW format of tensors after the ToTensor() function is used in torch 2.10, but instead remains in the HWC format? When I switch to version 2.9.1, the issue is resolved, and the code r… --- ## [Use Tensorboard Writer to track multiple model graph at once](https://discuss.pytorch.org/t/use-tensorboard-writer-to-track-multiple-model-graph-at-once/111045)
**Author:** [@CaoHoangTung](https://discuss.pytorch.org/u/CaoHoangTung)\ **Replies:** 1\ **Last updated:** [January 12, 2026, 1:35pm UTC](https://discuss.pytorch.org/t/use-tensorboard-writer-to-track-multiple-model-graph-at-once/111045 "2026-01-12T13:35:00Z")
I have a model which was constructed by 2 sub-models. I tried to log the 2 models graph structure to Tensorboard using writer.add\_graph(model1, input1) writer.add\_graph(model2, input2) I want to view the structure of b… --- ## [How am i supposed to code an axonal delay in my SNN?](https://discuss.pytorch.org/t/how-am-i-supposed-to-code-an-axonal-delay-in-my-snn/223163)
**Author:** [@Photon1](https://discuss.pytorch.org/u/Photon1)\ **Replies:** 0\ **Last updated:** [September 18, 2025, 7:15am UTC](https://discuss.pytorch.org/t/how-am-i-supposed-to-code-an-axonal-delay-in-my-snn/223163 "2025-09-18T07:15:54Z")
Hi, so i am trying to make work a convolusionnal SNN with axonal delay, with pytorch. But i don’t find any info on internet about how to code an axonal delay. Here i am trying to just make a kind of circular buffer. B… --- ## [No download option for histogram visualizations in TensorBoard (when using torch-pruning)](https://discuss.pytorch.org/t/no-download-option-for-histogram-visualizations-in-tensorboard-when-using-torch-pruning/222479)
**Author:** [@SadMoon](https://discuss.pytorch.org/u/SadMoon)\ **Replies:** 0\ **Last updated:** [August 19, 2025, 12:08pm UTC](https://discuss.pytorch.org/t/no-download-option-for-histogram-visualizations-in-tensorboard-when-using-torch-pruning/222479 "2025-08-19T12:08:54Z")
When using the torch-pruning library with TensorBoard for histogram visualization, I noticed there’s no direct download option for histograms in the TensorBoard interface. Currently, I have to rely on taking screenshots … --- ## [Tensorboard add\_graph (actually jit.trace) does not work with None inputs](https://discuss.pytorch.org/t/tensorboard-add-graph-actually-jit-trace-does-not-work-with-none-inputs/219863)
**Author:** [@turbotimon](https://discuss.pytorch.org/u/turbotimon)\ **Replies:** 1\ **Last updated:** [May 8, 2025, 9:52am UTC](https://discuss.pytorch.org/t/tensorboard-add-graph-actually-jit-trace-does-not-work-with-none-inputs/219863 "2025-05-08T09:52:55Z")
Hi, I have a problem where a forward method takes optional inputs: def forward(self, x\_in): x1, x2 = x\_in if x2 is None: ... so when I want to add\_graph for a case without the x2 provided… writer.add\_graph… --- ## [How to embed pytorch profiler to verl?](https://discuss.pytorch.org/t/how-to-embed-pytorch-profiler-to-verl/218916)
**Author:** [@gogineni\_kailash](https://discuss.pytorch.org/u/gogineni_kailash)\ **Replies:** 0\ **Last updated:** [April 9, 2025, 6:38pm UTC](https://discuss.pytorch.org/t/how-to-embed-pytorch-profiler-to-verl/218916 "2025-04-09T18:38:58Z")
This is my code base and I want to add pytorch profiler to the fit() function. Can anyone let me know how to embed the pytorch profiler? I also raised the issue in the verl github, but was not resolved. Any help or in… --- ## [How to export tensorboard data to .csv file](https://discuss.pytorch.org/t/how-to-export-tensorboard-data-to-csv-file/205004)
**Author:** [@AlphaBetaGamma96](https://discuss.pytorch.org/u/AlphaBetaGamma96)\ **Replies:** 0\ **Last updated:** [June 20, 2024, 3:56pm UTC](https://discuss.pytorch.org/t/how-to-export-tensorboard-data-to-csv-file/205004 "2024-06-20T15:56:12Z")
Hi All, I’m going through the tutorial of tensorboard (tutorial here: torch.utils.tensorboard — PyTorch 2.3 documentation), but I can’t see a way to export the data from a log to say .csv to allow post-processing to tak… --- ## [Hparams logs not found by tensorboard](https://discuss.pytorch.org/t/hparams-logs-not-found-by-tensorboard/201274)
**Author:** [@app31t](https://discuss.pytorch.org/u/app31t)\ **Replies:** 1\ **Last updated:** [April 22, 2024, 1:21pm UTC](https://discuss.pytorch.org/t/hparams-logs-not-found-by-tensorboard/201274 "2024-04-22T13:21:42Z")
I have been logging my metrics to tensorboard as scalars but wanted organise everything better by also logging hparams. However, tensorboard somehow can‘t display the logs. When I use writer.add\_hparams, the hparams car… --- ## [RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 18 but got size 17 for tensor number 1 in the list](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1-expected-size-18-but-got-size-17-for-tensor-number-1-in-the-list/194448)
**Author:** [@khaw235](https://discuss.pytorch.org/u/khaw235)\ **Replies:** 0\ **Last updated:** [December 27, 2023, 7:35pm UTC](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1-expected-size-18-but-got-size-17-for-tensor-number-1-in-the-list/194448 "2023-12-27T19:35:50Z")
Hi there, I am new to Pytorch and using the pre-trained model for my Django App which restores broken images. I took the help from the code given in this repository. My ML Model Class’s code goes like this: class Hour… --- ## [Model visualization for covariance between input variables](https://discuss.pytorch.org/t/model-visualization-for-covariance-between-input-variables/194242)
**Author:** [@hawkiyc](https://discuss.pytorch.org/u/hawkiyc)\ **Replies:** 0\ **Last updated:** [December 24, 2023, 9:03am UTC](https://discuss.pytorch.org/t/model-visualization-for-covariance-between-input-variables/194242 "2023-12-24T09:03:49Z")
Hi, I have built a VAE model dealing with tabular oceanographical data. My question is, what shall I do to visualize model weight and clarify the covariance relationship between input variables, as shown in the followi… --- ## [RuntimeError: Sizes of tensors must match except in dimension 1](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1/140651)
**Author:** [@Samah\_Abu\_saleem](https://discuss.pytorch.org/u/Samah_Abu_saleem)\ **Replies:** 13\ **Last updated:** [November 23, 2023, 1:45pm UTC](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1/140651 "2023-11-23T13:45:05Z")
I am trying to train GCN model on my custom dataset and I have resized all the values but I am getting error: RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 7 but got size 14515200 for t… --- ## [RuntimeError: Sizes of tensors must match except in dimension 1. Got 0 and 40 (The offending index is 0)](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1-got-0-and-40-the-offending-index-is-0/190989)
**Author:** [@yilin\_zhang](https://discuss.pytorch.org/u/yilin_zhang)\ **Replies:** 0\ **Last updated:** [October 31, 2023, 8:12am UTC](https://discuss.pytorch.org/t/runtimeerror-sizes-of-tensors-must-match-except-in-dimension-1-got-0-and-40-the-offending-index-is-0/190989 "2023-10-31T08:12:56Z")
I am trying to train my model on my custom dataset, but I am getting error: RuntimeError: Sizes of tensors must match except in dimension 1. Got 0 and 40 (The offending index is 0) This is my wrong code: import json … --- ## [Writer.add\_graph(model)](https://discuss.pytorch.org/t/writer-add-graph-model/153097)
**Author:** [@Rakshith\_V](https://discuss.pytorch.org/u/Rakshith_V)\ **Replies:** 2\ **Last updated:** [October 8, 2023, 3:21pm UTC](https://discuss.pytorch.org/t/writer-add-graph-model/153097 "2023-10-08T15:21:47Z")
I’m trying to pass the model to tensorboard. Model takes multiple inputs such as input\_ids,attention,etc. So, how exactly I should use writer.add\_graph in my case to make it work --- ## [How to add graphs to hparams in tensorboard?](https://discuss.pytorch.org/t/how-to-add-graphs-to-hparams-in-tensorboard/109349)
**Author:** [@aledjackson](https://discuss.pytorch.org/u/aledjackson)\ **Replies:** 3\ **Last updated:** [March 5, 2023, 7:47am UTC](https://discuss.pytorch.org/t/how-to-add-graphs-to-hparams-in-tensorboard/109349 "2023-03-05T07:47:27Z")
I’m trying to use tensorboard with pytorch and cannot find any clear documentation as to how to add graphs to the hparams dashboard. The image below shows what I want however I can only add singular values for the me… --- ## [RuntimeError: DataLoader worker (pid(s) 48916, 48917) exited unexpectedly](https://discuss.pytorch.org/t/runtimeerror-dataloader-worker-pid-s-48916-48917-exited-unexpectedly/180356)
**Author:** [@Killpit](https://discuss.pytorch.org/u/Killpit)\ **Replies:** 0\ **Last updated:** [May 20, 2023, 11:14am UTC](https://discuss.pytorch.org/t/runtimeerror-dataloader-worker-pid-s-48916-48917-exited-unexpectedly/180356 "2023-05-20T11:14:48Z")
I have encountered the error when I try to run the tutorial with tensorboard. Initially, it told me that tensorboard was not found and I installed tensorboard along with TensorFlow. Even though I had not encountered an i… --- ## [Summarywriter.add\_scalar() logs issue](https://discuss.pytorch.org/t/summarywriter-add-scalar-logs-issue/179495)
**Author:** [@viduzz84](https://discuss.pytorch.org/u/viduzz84)\ **Replies:** 0\ **Last updated:** [May 9, 2023, 10:23am UTC](https://discuss.pytorch.org/t/summarywriter-add-scalar-logs-issue/179495 "2023-05-09T10:23:30Z")
Hi, I experienced a problem in the behavior of tensorboard, when recording a scalar value with summarywriter.add\_scalar() In my experiment I use a set of loss criteria, whose values I record at each epoch, with add\_scal… --- ## [Error about tensorboard version](https://discuss.pytorch.org/t/error-about-tensorboard-version/179026)
**Author:** [@morteza\_mahvelati](https://discuss.pytorch.org/u/morteza_mahvelati)\ **Replies:** 1\ **Last updated:** [May 3, 2023, 3:23pm UTC](https://discuss.pytorch.org/t/error-about-tensorboard-version/179026 "2023-05-03T15:23:48Z")
I run my model without any problem, but newly i got an error :‘’ TensorBoard logging requires TensorBoard version 1.15 or above." i did some possible solutions such as upgrading tf, td, torch, … but the error has not go… --- ## [Tensorboard add\_histogram](https://discuss.pytorch.org/t/tensorboard-add-histogram/175025)
**Author:** [@Rakshith\_V](https://discuss.pytorch.org/u/Rakshith_V)\ **Replies:** 1\ **Last updated:** [March 16, 2023, 6:26pm UTC](https://discuss.pytorch.org/t/tensorboard-add-histogram/175025 "2023-03-16T18:26:27Z")
In the bin option of tensor-board’s add\_histogram method. I see the options such as “tensorflow,auto,fd”. Can someone how these options work ? (Except auto) --- ## [Converting tensorflow model to pytorch!](https://discuss.pytorch.org/t/converting-tensorflow-model-to-pytorch/170968)
**Author:** [@Nicolas\_Isla](https://discuss.pytorch.org/u/Nicolas_Isla)\ **Replies:** 1\ **Last updated:** [March 5, 2023, 9:16am UTC](https://discuss.pytorch.org/t/converting-tensorflow-model-to-pytorch/170968 "2023-03-05T09:16:29Z")
Greetings, My data consists of time-series samples with 925 steps, each containing 2 features. In other words, my data is shaped as (samples, steps, features) The model I’m currently implementing works in TensorFlow, b… --- ## [How can I connect previous logg file into new ones in tensorboard](https://discuss.pytorch.org/t/how-can-i-connect-previous-logg-file-into-new-ones-in-tensorboard/169321)
**Author:** [@test\_project](https://discuss.pytorch.org/u/test_project)\ **Replies:** 0\ **Last updated:** [December 31, 2022, 2:15am UTC](https://discuss.pytorch.org/t/how-can-i-connect-previous-logg-file-into-new-ones-in-tensorboard/169321 "2022-12-31T02:15:30Z")
I trained my model for 50 epochs. Now I want to continue training for other 15 epochs. so I start training with model.load\_from\_checkpoint("path") but I don’t know how to show continuation of logs with tensorboard I … --- ## [How do my transforms affect my model?](https://discuss.pytorch.org/t/how-do-my-transforms-affect-my-model/167292)
**Author:** [@gifted\_spence](https://discuss.pytorch.org/u/gifted_spence)\ **Replies:** 1\ **Last updated:** [December 1, 2022, 6:39pm UTC](https://discuss.pytorch.org/t/how-do-my-transforms-affect-my-model/167292 "2022-12-01T18:39:05Z")
Hey, I don’t understand why my transformations lead to my model failing to run (RuntimeError: mat1 and mat2 shapes cannot be multiplied (4x44944 and 400x120)). Generally, I stick to the documentation of training a class… --- ## [Making subplots using tensorboardX for pytorch](https://discuss.pytorch.org/t/making-subplots-using-tensorboardx-for-pytorch/167238)
**Author:** [@ZSheikhb](https://discuss.pytorch.org/u/ZSheikhb)\ **Replies:** 0\ **Last updated:** [November 30, 2022, 4:12pm UTC](https://discuss.pytorch.org/t/making-subplots-using-tensorboardx-for-pytorch/167238 "2022-11-30T16:12:40Z")
I’d like to make a set of scatter plots from a dimension of 3D tensor in the form of a combination of subplots inside tensorboard. The tensors have this shape like torch.Size(\[20,31,11\]). Here is my current attempt imp… --- ## [Tensorboard: graph view: how to show Op attributes?](https://discuss.pytorch.org/t/tensorboard-graph-view-how-to-show-op-attributes/167012)
**Author:** [@cee1](https://discuss.pytorch.org/u/cee1)\ **Replies:** 0\ **Last updated:** [November 27, 2022, 11:12am UTC](https://discuss.pytorch.org/t/tensorboard-graph-view-how-to-show-op-attributes/167012 "2022-11-27T11:12:15Z")
Hi all, With torch.utils.tensorboard, a model can be viewed in browser via SummaryWriter.add\_graph() Though, I’m unable to view the attributes of an Op e.g. kernel size, stride of a Conv2D: I guess these attributes… --- ## [AttributeError: 'GlobalStorage' object has no attribute 'float'](https://discuss.pytorch.org/t/attributeerror-globalstorage-object-has-no-attribute-float/165968)
**Author:** [@keerthi](https://discuss.pytorch.org/u/keerthi)\ **Replies:** 0\ **Last updated:** [November 14, 2022, 10:24pm UTC](https://discuss.pytorch.org/t/attributeerror-globalstorage-object-has-no-attribute-float/165968 "2022-11-14T22:24:27Z")
Hello, I am trying to run my GCN on image dataset. I have used dataloader and this is the following output of the data loader Below is my code for my model and my training of my data: class GCNN(torch.nn.Module): d… --- ## [Tensorboard is always showing old data](https://discuss.pytorch.org/t/tensorboard-is-always-showing-old-data/161169)
**Author:** [@Coca\_Cola12](https://discuss.pytorch.org/u/Coca_Cola12)\ **Replies:** 5\ **Last updated:** [September 12, 2022, 9:13pm UTC](https://discuss.pytorch.org/t/tensorboard-is-always-showing-old-data/161169 "2022-09-12T21:13:24Z")
I have a code that’s supposed to show a simple line on Tensorboard. import torch.utils.tensorboard as tb import tempfile log\_dir = tempfile.mkdtemp() print(log\_dir) %tensorboard --logdir {log\_dir} logger = tb.SummaryWr… --- ## [PyTorch profiler CUDA results not shown](https://discuss.pytorch.org/t/pytorch-profiler-cuda-results-not-shown/147127)
**Author:** [@Pramesh\_Gautam1](https://discuss.pytorch.org/u/Pramesh_Gautam1)\ **Replies:** 1\ **Last updated:** [August 19, 2022, 7:53am UTC](https://discuss.pytorch.org/t/pytorch-profiler-cuda-results-not-shown/147127 "2022-08-19T07:53:24Z")
I’ve been using PyTorch profiler and the results are attached here. The chart only shows DataLoader, CPU Exec and Other. Tensorboard chart is not showing GPU time. I’ve used activities=\[torch.profiler.ProfilerActivity.CU… --- ## [Relationship between tensorboard and Open3D?](https://discuss.pytorch.org/t/relationship-between-tensorboard-and-open3d/156225)
**Author:** [@jiapei-nexera](https://discuss.pytorch.org/u/jiapei-nexera)\ **Replies:** 1\ **Last updated:** [July 9, 2022, 12:47am UTC](https://discuss.pytorch.org/t/relationship-between-tensorboard-and-open3d/156225 "2022-07-09T00:47:06Z")
It’s so unbelievably crazy that when I built tensorboard from source, it requested from open3d.visualization import webrtc\_server, which seems to me have NOTHING to do with tensorboard … Can anybody help to explain why … --- ## [RuntimeError: expected scalar type Byte but found Float](https://discuss.pytorch.org/t/runtimeerror-expected-scalar-type-byte-but-found-float/141897)
**Author:** [@Samah\_Abu\_saleem](https://discuss.pytorch.org/u/Samah_Abu_saleem)\ **Replies:** 6\ **Last updated:** [May 22, 2022, 4:57pm UTC](https://discuss.pytorch.org/t/runtimeerror-expected-scalar-type-byte-but-found-float/141897 "2022-05-22T16:57:13Z")
I want to implement GCN on my data. the data looks like this: Data(pos=\[4, 2\], x=\[4, 1080, 1920, 3\], y=\[4\], edge\_index=\[2, 12\], edge\_attr=\[12, 1\]) my code: # Set optimizer (adam) optimizer = torch.optim.Adam(model.pa… --- ## [Time measurement of each layer?](https://discuss.pytorch.org/t/time-measurement-of-each-layer/145810)
**Author:** [@koala](https://discuss.pytorch.org/u/koala)\ **Replies:** 2\ **Last updated:** [May 7, 2022, 10:32am UTC](https://discuss.pytorch.org/t/time-measurement-of-each-layer/145810 "2022-05-07T10:32:39Z")
Hi, Can anybody tell me how to use tensorboard module in pytorch to measuring inference time of each layer? I appreciate any reply. Thank you so much~ [Next page](https://discuss.pytorch.org/c/tensorboard/28.md?page=1)