issue-tracking
issue-tracking copied to clipboard
comet-ml
Questions, Help, and Issues for Comet ML
When in last page of archived experiments, deleting all experiments shown results in `This project has no archived experiments.` even though there are more archived experiments and the counter shows...
## Describe the Bug The application runs out of memory and is killed attempting to `log_histogram_3d` with a Pytorch tensor on the GPU. ## Expected behavior Either of the following...
Currently when we run a tensorflow job and a ``` tf.summary.image("a_name", png_as_tensor, step=1) tf.summary.image("a_name", png_as_tensor, step=2) ``` we get a proper grouping by name in tensorboard, but in comet the...
I used ``` comet_ml.Experiment( project_name="foo", workspace="username", log_code=False, log_git_metadata=True, log_git_patch=True ) ``` and my experiment reports says "No commit available". When I change it to ``` comet_ml.Experiment( project_name="foo", workspace="username", log_code=True, #...
The deletion of an experiment in the online user interface does not delete the corresponding artifacts. This is really cumbersome because it requires the users to find the artifacts without...
Hello, As you can see in the image below, I have several curves under a particular experiment. However, none of them are being shown in Logged Curves.  For instance,...
Hi, I'm using [set_epoch](https://www.comet.ml/docs/python-sdk/Experiment/#experimentset_epoch) as in ``` for epoch in range(self.hyper_params['num_epochs']): self.comet.set_epoch(epoch) with self.comet.train(): logger.info(f"Training (epoch #{epoch}) ...") self.model.train() epoch_loss = self.train_one_epoch() / len(self.data_train_loader) logger.info( f"Training (epoch #{epoch}) completed with...
When selecting archived experiments, currently we can select at most the ones shown in the screen (max 25 since this is not customizable), but it would be useful to allow...
I create an experiment with `experiment = Experiment( api_key=args.comet_api_key, workspace=args.comet_workspace, project_name=args.project_name, auto_output_logging="simple", auto_metric_logging=False, parse_args=False, disabled=args.disable_comet or rank != 0)` and see system metrics tracked. My script is interrupted and my...
The confusion matrix is awesome! Im trying to plot also some interactive 2d visualization such as in https://umap-learn.readthedocs.io/en/latest/basic_usage.html  Any suggestions how to do it?
