[SDK] Add docstring for Train API #2075
Conversation
Signed-off-by: Andrey Velichkevich <[email protected]>
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: andreyvelich The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
Pull Request Test Coverage Report for Build 8760737620Details
💛 - Coveralls |
Signed-off-by: Andrey Velichkevich <[email protected]>
| ) | ||
|
|
||
| print( | ||
| "Thank you for using `train` API for LLMs fine-tuning. This feature is in alpha stage " |
There was a problem hiding this comment.
What is the definition of the alpha stage in SDK?
If we say that this feature is still alpha stage, defininig criteria to graduate them to beta and GA would be better.
There was a problem hiding this comment.
Yes, I know that @StefanoFioravanzo is working on this as part of this effort: kubeflow/community#709. When we have some time we can discuss what is the graduation steps for our new features.
Also, we do have criteria for Kubeflow components here: https://www.kubeflow.org/docs/started/support/#application-status, but it is not related to the components features.
There was a problem hiding this comment.
But, both (kubeflow/community#709 and https://www.kubeflow.org/docs/started/support/#application-status) seem to be for components, not SDK.
This PR seems to introduce a dedicated graduation cycle for SDK separated from the application above, right?
There was a problem hiding this comment.
@tenzen-y If you check @StefanoFioravanzo 's issue he also says this:
With "component" I not only mean an entire project/repo but also significant features within those projects (think about the new Fine Tune APIs for LLM).
So we also consider on how we should define graduation criteria for our new big features.
There was a problem hiding this comment.
Oh, I have missed that sentence. Thank you!
There was a problem hiding this comment.
Yes we need to resume the component/feature maturity discussion. Thank you @andreyvelich for adding that warning!
|
/hold cancel |
* [SDK] Add docstring for Train API Signed-off-by: Andrey Velichkevich <[email protected]> * Update a few args Signed-off-by: Andrey Velichkevich <[email protected]> --------- Signed-off-by: Andrey Velichkevich <[email protected]> Signed-off-by: Weiyu Yen <[email protected]>
* [SDK] Add docstring for Train API Signed-off-by: Andrey Velichkevich <[email protected]> * Update a few args Signed-off-by: Andrey Velichkevich <[email protected]> --------- Signed-off-by: Andrey Velichkevich <[email protected]> Signed-off-by: Weiyu Yen <[email protected]>
* [SDK] Add docstring for Train API Signed-off-by: Andrey Velichkevich <[email protected]> * Update a few args Signed-off-by: Andrey Velichkevich <[email protected]> --------- Signed-off-by: Andrey Velichkevich <[email protected]>
* [SDK] Add docstring for Train API Signed-off-by: Andrey Velichkevich <[email protected]> * Update a few args Signed-off-by: Andrey Velichkevich <[email protected]> --------- Signed-off-by: Andrey Velichkevich <[email protected]>
Related: #2013.
I added valid docstring for
trainAPI and addedprintmessage for users to provide a feedback.Also, I renamed:
HuggingFaceTrainParams -> HuggingFaceTrainerParams
train_parameters -> trainer_parameters
We need to update SDK examples once we merge this change.
/assign @StefanoFioravanzo @kubeflow/wg-training-leads @deepanker13
/hold for review