-
Notifications
You must be signed in to change notification settings - Fork 281
Integrate new-developed pruning API with old one. #226
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: YIYANGCAI <[email protected]>
|
this title is not suitable as a commit message. you don't need mention merge things. you can mentioned new pruning algo or something else |
Signed-off-by: wenhuach <dd>
...les/pytorch/nlp/huggingface_models/question-answering/pruning/pytorch_pruner/eager/README.md
Outdated
Show resolved
Hide resolved
| end_step=int(total_iterations)) ##iterative | ||
| else: | ||
| pruner.update_items_for_all_pruners(start_step=total_iterations+1, end_step=total_iterations+1) ## remove the pruner | ||
| pruner.update_items_for_all_pruners(start_step=total_iterations + 1, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
shall this func get exposed to external user? is it possible to hide this part?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Some configs like start_step and end_step are affected by the training hyperparameters, which users should usually tune. This function eases users to change these configs. I suggest keeping it, but of course, removing it does not affect the functionality
Signed-off-by: YIYANGCAI <[email protected]> Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: Zhang, Weiwei1 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
|
Do we need this PR: #144 |
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
…al-compressor into cyy/pruning-api-merge
Signed-off-by: wenhuach21 <[email protected]>
|
@wenhuach21 there is conflict to be solved. and I saw the neural_compressor/conf/config.py is modified? this will impact backward compatibility. you should only modify neural_compressor/config.py which is newly introduce for inc 2.0. |
|
as I mentioned before, we have 3 sets of API due to historical reason.
|
Signed-off-by: YIYANGCAI <[email protected]>
Signed-off-by: Lu, Yintong <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: wenhuach21 <[email protected]>
Signed-off-by: Lu, Yintong <[email protected]>
…al-compressor into cyy/pruning-api-merge
Signed-off-by: YIYANGCAI <[email protected]>
Signed-off-by: Lu, Yintong <[email protected]>
Signed-off-by: YIYANGCAI [email protected]
Type of Change
Merge two pruning APIs to one.
API changed
Description
will be linked to PR#144