ClearML/clearml-docs
1
0
Fork 0
mirror of https://github.com/clearml/clearml-docs synced 2025-06-26 18:17:44 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge branch 'main' of https://github.com/allegroai/clearml-docs into edits_2

Browse Source
This commit is contained in:
revital
2025-03-11 17:44:52 +02:00
parent 9e275de492 b987ca5113
commit f6cff8e460
145 changed files with 250 additions and 137 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
docs/integrations/catboost.md
View File

@@ -31,7 +31,8 @@ You can view all the task details in the [WebApp](../webapp/webapp_overview.md).
See an example of CatBoost and ClearML in action [here](../guides/frameworks/catboost/catboost.md).
![Task scalars](../img/examples_catboost_scalars.png)
![Task scalars](../img/examples_catboost_scalars.png#light-mode-only)
![Task scalars](../img/examples_catboost_scalars_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your CatBoost script, it captures models, and

3
docs/integrations/click.md
View File

@@ -42,7 +42,8 @@ if __name__ == '__main__':
When this code is executed, ClearML logs your command-line arguments, which you can view in the
[WebApp](../webapp/webapp_overview.md), in the task's **Configuration > Hyperparameters > Args** section.
![click configuration](../img/integrations_click_configs.png)
![click configuration](../img/integrations_click_configs.png#light-mode-only)
![click configuration](../img/integrations_click_configs_dark.png#dark-mode-only)
In the UI, you can clone the task multiple times and set the clones' parameter values for execution by the [ClearML Agent](../clearml_agent.md).
When the clone is executed, the executing agent will use the new parameter values as if set by the command-line.

3
docs/integrations/fastai.md
View File

@@ -30,7 +30,8 @@ You can view all the task details in the [WebApp](../webapp/webapp_overview.md).
See an example of `fastai` and ClearML in action [here](../guides/frameworks/fastai/fastai_with_tensorboard.md).
![Task scalars](../img/examples_reporting_fastai_01.png)
![Task scalars](../img/examples_reporting_fastai_01.png#light-mode-only)
![Task scalars](../img/examples_reporting_fastai_01_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your `fastai` script, it captures models and

3
docs/integrations/hydra.md
View File

@@ -22,7 +22,8 @@ task = Task.init(task_name="<task_name>", project_name="<project_name>")
ClearML logs the OmegaConf as a blob and can be viewed in the
[WebApp](../webapp/webapp_overview.md), in the task's **CONFIGURATION > CONFIGURATION OBJECTS > OmegaConf** section.
![Hydra configuration](../img/integrations_hydra_configs.png)
![Hydra configuration](../img/integrations_hydra_configs.png#light-mode-only)
![Hydra configuration](../img/integrations_hydra_configs_dark.png#dark-mode-only)
## Modifying Hydra Values

1
docs/integrations/integrations.md
View File

@@ -8,6 +8,7 @@ ClearML seamlessly integrates with a wide range of popular machine learning fram
* [Keras](keras.md)
* [YOLO v5](yolov5.md)
* [YOLO v8](yolov8.md)
* [Hugging Face Transformers](transformers.md)
* [MMEngine](mmengine.md)
* [MMCV](mmcv.md)
* [MONAI](monai.md)

9
docs/integrations/keras_tuner.md
View File

@@ -53,16 +53,19 @@ You can view all the task details in the [WebApp](../webapp/webapp_exp_track_vis
ClearML logs the scalars from training each network. They appear in the task's **SCALARS** tab in the Web UI.
![Optimization scalars](../img/integration_keras_tuner_06.png)
![Optimization scalars](../img/integration_keras_tuner_06.png#light-mode-only)
![Optimization scalars](../img/integration_keras_tuner_06_dark.png#dark-mode-only)
ClearML automatically logs the parameters of each task run in the hyperparameter search. They appear in tabular
form in the task's **PLOTS**.
![Optimization plot](../img/integration_keras_tuner_07.png)
![Optimization plot](../img/integration_keras_tuner_07.png#light-mode-only)
![Optimization plot](../img/integration_keras_tuner_07_dark.png#dark-mode-only)
ClearML automatically stores the output model. It appears in the task's **ARTIFACTS** **>** **Output Model**.
![output model](../img/integration_keras_tuner_03.png)
![output model](../img/integration_keras_tuner_03.png#light-mode-only)
![output model](../img/integration_keras_tuner_03_dark.png#dark-mode-only)
## Example

3
docs/integrations/pytorch.md
View File

@@ -28,7 +28,8 @@ And that's it! This creates a [ClearML Task](../fundamentals/task.md) which capt
You can view all the task details in the [WebApp](../webapp/webapp_overview.md).
![Pytorch webapp](../img/examples_pytorch_distributed_example_08.png)
![Pytorch webapp](../img/examples_pytorch_distributed_example_08.png#light-mode-only)
![Pytorch webapp](../img/examples_pytorch_distributed_example_08_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your PyTorch script, it captures PyTorch models. But, you may want to have

6
docs/integrations/tensorboard.md
View File

@@ -22,9 +22,11 @@ uncommitted code, Python environment, your TensorBoard metrics, plots, images, a
View the TensorBoard outputs in the [WebApp](../webapp/webapp_overview.md), in the task's page.
![TensorBoard WebApp scalars](../img/examples_pytorch_tensorboard_07.png)
![TensorBoard WebApp scalars](../img/examples_pytorch_tensorboard_07.png#light-mode-only)
![TensorBoard WebApp scalars](../img/examples_pytorch_tensorboard_07_dark.png#dark-mode-only)
![Tensorboard WebApp debug samples](../img/examples_tensorboard_toy_pytorch_02.png)
![Tensorboard WebApp debug samples](../img/examples_tensorboard_toy_pytorch_02.png#light-mode-only)
![Tensorboard WebApp debug samples](../img/examples_tensorboard_toy_pytorch_02_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your script, it captures all of your TensorBoard plots, images, and metrics.

3
docs/integrations/tensorboardx.md
View File

@@ -22,7 +22,8 @@ uncommitted code, Python environment, your TensorboardX metrics, plots, images,
View the TensorboardX outputs in the [WebApp](../webapp/webapp_overview.md), in the task's page.
![TensorboardX WebApp scalars](../img/examples_pytorch_tensorboardx_03.png)
![TensorboardX WebApp scalars](../img/examples_pytorch_tensorboardx_03.png#light-mode-only)
![TensorboardX WebApp scalars](../img/examples_pytorch_tensorboardx_03_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your script, it captures all of your TensorboardX plots, images, metrics, videos, and text.

25
docs/integrations/transformers.md
View File

@@ -77,16 +77,29 @@ cloud of your choice (AWS, GCP, Azure) and automatically deploy ClearML agents:
and shuts down instances as needed, according to a resource budget that you set.
### Cloning, Editing, and Enqueuing
### Reproducing Tasks
![Cloning, editing, enqueuing gif](../img/gif/integrations_yolov5.gif#light-mode-only)
![Cloning, editing, enqueuing gif](../img/gif/integrations_yolov5_dark.gif#dark-mode-only)
Use ClearML's web interface to edit task details, like configuration parameters or input models, then execute the task
with the new configuration on a remote machine:
* Clone the task
* Edit the hyperparameters and/or other details
* Enqueue the task
Use ClearML's web interface to reproduce tasks and edit their details, like hyperparameters or input models, then execute the tasks
with the new configuration on a remote machine.
When ClearML is integrated into a script, it captures and stores configurations, such as hyperparameters
and model settings. When executing a task, the ClearML Agent will, by default, override runtime configuration values
(such as hyperparameters and environment variables) with the values specified in the task.
However, for tasks using Transformers, the default behavior is different. By default, Transformers tasks ignore UI
overrides and use execution-time parameters (such as environment variables). This is done to prevent potential issues
with environment-specific settings when running tasks on different machines.
**To rerun a task with modified configuration:**
1. Clone the task
1. Edit the hyperparameters and/or other details.
1. In the **CONFIGURATION > HYPERPARAMETERS > Transformers** section, set both `_ignore_hparams_ui_overrides_` and `_ignore_model_config_ui_overrides_`
to `False` . This allows the task to use the new hyperparameter and model
configuration values respectively during execution.
1. Enqueue the task
The ClearML Agent executing the task will use the new values to [override any hard coded values](../clearml_agent.md).

3
docs/integrations/xgboost.md
View File

@@ -51,7 +51,8 @@ except ImportError:
You can view all the task details in the [WebApp](../webapp/webapp_overview.md).
![Task scalars](../img/examples_xgboost_metric_scalars.png)
![Task scalars](../img/examples_xgboost_metric_scalars.png#light-mode-only)
![Task scalars](../img/examples_xgboost_metric_scalars_dark.png#dark-mode-only)
## Automatic Logging Control
By default, when ClearML is integrated into your XGBoost script, it captures models, and