Update docs (#520)

docs/clearml_sdk/model_sdk.md

@@ -101,7 +101,8 @@ input_model = InputModel.import_model(
 task.connect(input_model)
 ```
 
-## Querying Models
+## Accessing Models
+### Querying Models
 Retrieve a list of model objects by querying the system by model names, projects, tags, and more, using the 
 [`Model.query_models`](../references/sdk/model_model.md#modelquery_models) and / or 
 the [`InputModel.query_models`](../references/sdk/model_inputmodel.md#inputmodelquery_models) class methods. These 
@@ -127,7 +128,7 @@ model_list = Model.query_models(
 )
 ```
 
-### Tag Filters
+#### Tag Filters
 The `tags` field supports advanced queries through combining tag names and operators into a list. 
 
 The supported operators are: 
@@ -177,6 +178,31 @@ The default operator for a query is `or`, unless `and` is placed at the beginnin
   )
   ```
 
+### Retrieving Models 
+Retrieve a local copy of a ClearML model through a `Model`/`InputModel` object's [`get_local_copy()`](../references/sdk/model_outputmodel.md#get_local_copy). 
+The method returns a path to a cached local copy of the model. In the case that the model is already cached, you can set 
+`force_download` to `True` in order to download a fresh version.
+
+## Logging Metrics and Plots
+
+Use the following methods to explicitly log additional information to your models. 
+These methods can be used on `Model`, `InputModel`, and/or `OutputModel` objects:
+* Scalars 
+  * Scalar series plots - [`report_scalar`](../references/sdk/model_outputmodel.md#report_scalar) 
+  * Single metric values - [`report_single_value`](../references/sdk/model_outputmodel.md#report_single_value)
+* Plots
+  * 2d plots
+    * Histogram - [`report_histogram`](../references/sdk/model_outputmodel.md#report_historgram)
+    * Vector as histogram plot - [`report_vector`](../references/sdk/model_outputmodel.md#report_vector)
+    * Table - [`report_table`](../references/sdk/model_outputmodel.md#report_table)
+    * Line plot - [`report_line_plot`](../references/sdk/model_outputmodel.md#report_line_plot)
+    * Scatter plot - [`report_scatter2d`](../references/sdk/model_outputmodel.md#report_scatter2d)
+    * Confusion matrix (heat map) - [`report_confusion_matrix`](../references/sdk/model_outputmodel.md#report_confusion_matrix) & [`report_matrix`](../references/sdk/model_outputmodel.md#report_matrix)
+  * 3d plots 
+    * Scatter plot - [`report_scatter3d`](../references/sdk/model_outputmodel.md#report_scatter3d) 
+    * Surface plot - [`report_surface`](../references/sdk/model_outputmodel.md#report_surface)
+
+  
 ## SDK Reference
 
 For information about all model methods, see the following SDK reference pages:

docs/clearml_sdk/task_sdk.md

@@ -409,6 +409,11 @@ folder, which can later be uploaded to the [ClearML Server](../deploying_clearml
 Before initializing a Task, use the [Task.set_offline](../references/sdk/task.md#taskset_offline) class method and set 
 the `offline_mode` argument to `True`. The method returns the Task ID and a path to the session folder. 
 
+:::caution 
+Notice that the `Task.set_offline` method only works with tasks created using `Task.init` and not with those created 
+using the `Task.create` method. 
+:::
+
 ```python
 from clearml import Task
 # Use the set_offline class method before initializing a Task
@@ -436,6 +441,17 @@ Upload the execution data that the Task captured offline to the ClearML Server u
 
   In the `session_folder_zip` argument, insert the path to the zip folder containing the session.
   
+  To upload the session from the same script that created it, first close the task then disable offline mode: 
+  
+  ```python
+  Task.set_offline(offline_mode=True)
+  task = Task.init(project_name="examples", task_name="my_task")
+  # task code
+  task.close()
+  Task.set_offline(False)
+  Task.import_offline_session(task.get_offline_mode_folder())
+  ```
+
   You can also use the offline task to update the execution of an existing previously executed task by providing the 
   previously executed task’s ID. To avoid overwriting metrics, you can specify the initial iteration offset with 
   `iteration_offset`.   

docs/guides/set_offline.md

@@ -11,6 +11,12 @@ local folder, which can be later uploaded to the [ClearML Server](../deploying_c
 Before initializing a Task, use the [Task.set_offline](../references/sdk/task.md#taskset_offline) class method and set the 
 `offline_mode` argument to `True`.
 
+:::caution 
+Notice that the `Task.set_offline` method only works with tasks created using `Task.init` and not with those created 
+using the `Task.create` method. 
+:::
+
+
 ```python
 from clearml import Task
 # Use the set_offline class method before initializing a Task
@@ -58,6 +64,17 @@ Upload the session's execution data that the Task captured offline to the ClearM
 
   In the `session_folder_zip` argument, insert the path to the zip folder containing the session.
   
+  To upload the session from the same script that created it, first close the task then disable offline mode: 
+  
+  ```python
+  Task.set_offline(offline_mode=True)
+  task = Task.init(project_name="examples", task_name="my_task")
+  # task code
+  task.close()
+  Task.set_offline(False)
+  Task.import_offline_session(task.get_offline_mode_folder())
+  ```
+    
   You can also use the offline task to update the execution of an existing previously executed task by providing the 
   previously executed task’s ID. To avoid overwriting metrics, you can specify the initial iteration offset with 
   `iteration_offset`.