Skip to content

Add nni_auto_tune example - #325

Merged
MaggieQi merged 25 commits into
microsoft:mainfrom
smallv0221:add-ann-auto-tune
Nov 1, 2022
Merged

MaggieQi merged 25 commits into
microsoft:mainfrom
smallv0221:add-ann-auto-tune

Conversation

@smallv0221

Copy link
Copy Markdown
Contributor

Add nni_auto_tune examples for sift and glove100.

@ghost

Deleted user (ghost) commented Jul 27, 2022 •

Copy link
Copy Markdown

CLA assistant check
All CLA requirements met.

Comment thread tools/nni-auto-tune/search_space.json Resolved
Comment thread tools/nni-auto-tune/config_aml.yml Resolved
Comment thread tools/nni-auto-tune/main.py Resolved
Comment thread tools/nni-auto-tune/plot.py Outdated
comparator = ((lambda xv, lx: xv > lx) if last_x < 0 else
(lambda xv, lx: xv < lx))
for i, (q, r, b, bp, sp) in enumerate(data):
if comparator(r, last_x):

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you explain the selection rule here?

@suiguoxin

Copy link
Copy Markdown
Contributor

How do you get the performance of hnswlib and scann?

MaggieQi
MaggieQi previously approved these changes Aug 31, 2022
Comment thread tools/nni-auto-tune/model.py Resolved
Comment thread tools/nni-auto-tune/main.py Resolved
Comment thread tools/nni-auto-tune/main.py Resolved
Comment thread tools/nni-auto-tune/main.py Resolved

@suiguoxin Guoxin (suiguoxin) left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please resolve data type issues for int8 & float64

Comment thread tools/nni-auto-tune/main.py Resolved
'int8': 'Int8',
'int16': 'Int16'
}[data_type]
self._sptag = SPTAG.AnnIndex(self._algo, self._data_type, X.shape[1])

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does capitalization of data_type matter?

@smallv0221

Copy link
Copy Markdown
Contributor Author

@microsoft-github-policy-service agree [company="{Microsoft}"]

@smallv0221 smallv0221 closed this Oct 20, 2022
@smallv0221 smallv0221 reopened this Oct 20, 2022
MaggieQi
MaggieQi previously approved these changes Oct 20, 2022

@MaggieQi MaggieQi left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

build ci

@MaggieQi
MaggieQi merged commit 636b3d6 into microsoft:main Nov 1, 2022
MaggieQi added a commit that referenced this pull request Nov 3, 2022
* Fixed bugs with accuracy fix for TPT build with recon vectors

* Accuracy fix working for GPU index build with PQ/OPQ enabled

* Fix bug with accuracy fix

* Trigger CI

* Fix low accuracy issue with GPU index build for int8/cosine configuration (#339)

* Fix bug with int8/cosine configuration, and enabled hardware optimization for this case.

* Trigger CI

Co-authored-by: diegocai <diegocai@microsoft.com>

* Add nni_auto_tune example (#325)

* add nni_auto_tune

* support other data format and add more result

* add result picture

* refine readme

* refine readme

* refine code style

* update readme and datareader

* add aml training config

* update config

* fix licence

* refactor for data type

* refactor for data type

* refactor for data type

* fix overflow on bruteforce

* fix compute metric by index

* add limits and preprocessing

* change code dir

Co-authored-by: Guoxin <suiguoxin@gmail.com>
Co-authored-by: MaggieQi <chenqi871025@gmail.com>

* Fixed bugs with accuracy fix for TPT build with recon vectors

* Accuracy fix working for GPU index build with PQ/OPQ enabled

* Fix bug with accuracy fix

* Trigger CI

Co-authored-by: diegocai <diegocai@microsoft.com>
Co-authored-by: smallv0221 <33639025+smallv0221@users.noreply.github.com>
Co-authored-by: Guoxin <suiguoxin@gmail.com>
Co-authored-by: MaggieQi <chenqi871025@gmail.com>
L1ray (L1-ray) pushed a commit to L1-ray/SPTAG that referenced this pull request May 9, 2026
* add nni_auto_tune

* support other data format and add more result

* add result picture

* refine readme

* refine readme

* refine code style

* update readme and datareader

* add aml training config

* update config

* fix licence

* refactor for data type

* refactor for data type

* refactor for data type

* fix overflow on bruteforce

* fix compute metric by index

* add limits and preprocessing

* change code dir

Co-authored-by: Guoxin <suiguoxin@gmail.com>
Co-authored-by: MaggieQi <chenqi871025@gmail.com>
L1ray (L1-ray) pushed a commit to L1-ray/SPTAG that referenced this pull request May 9, 2026
* Fixed bugs with accuracy fix for TPT build with recon vectors

* Accuracy fix working for GPU index build with PQ/OPQ enabled

* Fix bug with accuracy fix

* Trigger CI

* Fix low accuracy issue with GPU index build for int8/cosine configuration (microsoft#339)

* Fix bug with int8/cosine configuration, and enabled hardware optimization for this case.

* Trigger CI

Co-authored-by: diegocai <diegocai@microsoft.com>

* Add nni_auto_tune example (microsoft#325)

* add nni_auto_tune

* support other data format and add more result

* add result picture

* refine readme

* refine readme

* refine code style

* update readme and datareader

* add aml training config

* update config

* fix licence

* refactor for data type

* refactor for data type

* refactor for data type

* fix overflow on bruteforce

* fix compute metric by index

* add limits and preprocessing

* change code dir

Co-authored-by: Guoxin <suiguoxin@gmail.com>
Co-authored-by: MaggieQi <chenqi871025@gmail.com>

* Fixed bugs with accuracy fix for TPT build with recon vectors

* Accuracy fix working for GPU index build with PQ/OPQ enabled

* Fix bug with accuracy fix

* Trigger CI

Co-authored-by: diegocai <diegocai@microsoft.com>
Co-authored-by: smallv0221 <33639025+smallv0221@users.noreply.github.com>
Co-authored-by: Guoxin <suiguoxin@gmail.com>
Co-authored-by: MaggieQi <chenqi871025@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants