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

Small edits (#668)

Browse Source
This commit is contained in:
pollfly
2023-09-11 13:33:30 +03:00
committed by GitHub
parent dec2ff2e1e
commit d2dbd30bb4
34 changed files with 53 additions and 62 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
docs/guides/distributed/distributed_pytorch_example.md
View File

@@ -14,15 +14,15 @@ dataset), and reports (uploads) the following to the main Task:
* Scalars - Loss reported as a scalar during training in each Task in a subprocess.
* Hyperparameters - Hyperparameters created in each Task are added to the hyperparameters in the main Task.
Each Task in a subprocess references the main Task by calling [Task.current_task](../../references/sdk/task.md#taskcurrent_task), which always returns
Each Task in a subprocess references the main Task by calling [`Task.current_task()`](../../references/sdk/task.md#taskcurrent_task), which always returns
the main Task.
When the script runs, it creates an experiment named `test torch distributed` in the `examples` project.
## Artifacts
The example uploads a dictionary as an artifact in the main Task by calling the [Task.upload_artifact](../../references/sdk/task.md#upload_artifact)
method on [`Task.current_task`](../../references/sdk/task.md#taskcurrent_task) (the main Task). The dictionary contains the [`dist.rank`](https://pytorch.org/docs/stable/distributed.html#torch.distributed.get_rank)
The example uploads a dictionary as an artifact in the main Task by calling [`Task.upload_artifact()`](../../references/sdk/task.md#upload_artifact)
on [`Task.current_task`](../../references/sdk/task.md#taskcurrent_task) (the main Task). The dictionary contains the [`dist.rank`](https://pytorch.org/docs/stable/distributed.html#torch.distributed.get_rank)
of the subprocess, making each unique.
```python
@@ -38,8 +38,8 @@ All of these artifacts appear in the main Task under **ARTIFACTS** **>** **OTHER
## Scalars
Loss is reported to the main Task by calling the [Logger.report_scalar](../../references/sdk/logger.md#report_scalar)
method on `Task.current_task().get_logger`, which is the logger for the main Task. Since `Logger.report_scalar` is called
Loss is reported to the main Task by calling the [`Logger.report_scalar()`](../../references/sdk/logger.md#report_scalar)
on `Task.current_task().get_logger()`, which is the logger for the main Task. Since `Logger.report_scalar` is called
with the same title (`loss`), but a different series name (containing the subprocess' `rank`), all loss scalar series are
logged together.

4
docs/guides/distributed/subprocess_example.md
View File

@@ -5,7 +5,7 @@ title: Subprocess
The [subprocess_example.py](https://github.com/allegroai/clearml/blob/master/examples/distributed/subprocess_example.py)
script demonstrates multiple subprocesses interacting and reporting to a main Task. The following happens in the script:
* This script initializes a main Task and spawns subprocesses, each for an instances of that Task.
* Each Task in a subprocess references the main Task by calling [Task.current_task](../../references/sdk/task.md#taskcurrent_task),
* Each Task in a subprocess references the main Task by calling [`Task.current_task()`](../../references/sdk/task.md#taskcurrent_task),
which always returns the main Task.
* The Task in each subprocess reports the following to the main Task:
* Hyperparameters - Additional, different hyperparameters.
@@ -15,7 +15,7 @@ which always returns the main Task.
## Hyperparameters
ClearML automatically logs the command line options defined with `argparse`. A parameter dictionary is logged by
connecting it to the Task using a call to the [`Task.connect`](../../references/sdk/task.md#connect) method.
connecting it to the Task using [`Task.connect()`](../../references/sdk/task.md#connect).
```python
additional_parameters = {