ClearML/clearml-docs
1
0
Fork 0
mirror of https://github.com/clearml/clearml-docs synced 2025-01-31 14:37:18 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Add configuration sentence to FAQ (#335)

Browse Source
This commit is contained in:
pollfly 2022-09-19 10:20:23 +03:00 committed by GitHub
parent f2277de827
commit 125d5b62df
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
docs/faq.md
View File

@ -283,10 +283,12 @@ to reproduce. You can see uncommitted changes in the ClearML Web UI, in the EXEC
**I do not use argparse for hyperparameters. Do you have a solution?** <a id="dont-want-argparser"></a> **I do not use argparse for hyperparameters. Do you have a solution?** <a id="dont-want-argparser"></a>
Yes! ClearML also automatically captures and tracks command line parameters created using [click](https://click.palletsprojects.com/), Yes! ClearML provides multiple ways to configure your task and track your parameters!
In addition to argparse, ClearML also automatically captures and tracks command line parameters created using [click](https://click.palletsprojects.com/),
[Python Fire](https://github.com/google/python-fire), and/or [LightningCLI](https://pytorch-lightning.readthedocs.io/en/stable/api/pytorch_lightning.utilities.cli.html). [Python Fire](https://github.com/google/python-fire), and/or [LightningCLI](https://pytorch-lightning.readthedocs.io/en/stable/api/pytorch_lightning.utilities.cli.html).
Additionally, ClearML supports connecting hyperparameter dictionaries to experiments, using the [Task.connect](references/sdk/task.md#connect) method. ClearML also supports tracking code-level configuration dictionaries using the [Task.connect](references/sdk/task.md#connect) method.
For example, the code below connects hyperparameters (`learning_rate`, `batch_size`, `display_step`, For example, the code below connects hyperparameters (`learning_rate`, `batch_size`, `display_step`,
`model_path`, `n_hidden_1`, and `n_hidden_2`) to a task: `model_path`, `n_hidden_1`, and `n_hidden_2`) to a task:
@ -300,7 +302,7 @@ parameters_dict = { 'learning_rate': 0.001, 'batch_size': 100, 'display_step': 1
parameters_dict = Task.current_task().connect(parameters_dict) parameters_dict = Task.current_task().connect(parameters_dict)
``` ```
See more information about logging of hyperparameters [here](fundamentals/hyperparameters.md). See more task configuration options [here](fundamentals/hyperparameters.md).
<br/> <br/>