[Release_v2150] Update ReleaseNotes.md #3214

nikita-malininn · 2025-01-27T08:58:25Z

Changes

Added v2.15.0 template;

Reason for changes

Upcoming release;

Related tickets

161230;

For the contributors:

Please add your changes (as the commit to the branch) to the list according to the template and previous notes;
Do not add tests-related notes;
Provide the list of the PRs (for all your notes) in the comment for the discussion;

nikita-malininn · 2025-01-27T09:02:21Z

@alexsu52, @ljaljushkin, @l-bat, @nikita-savelyevv, @andreyanufr, @andrey-churkin, @daniil-lyakhov, @kshpv, @AlexanderDokuchaev, @anzr299 fill the document with your changes for the upcoming release, please.

ljaljushkin · 2025-01-27T14:16:48Z

no changes from my side

MaximProshin · 2025-01-28T08:16:43Z

@andrey-churkin , please add the deprecation note about create_compressed_model() in TF among the description of the related changes and the reference to the example.

MaximProshin · 2025-01-28T08:20:34Z

@l-bat , please update the list of new/updated notebooks with NNCF support. Here is a draft list from my side (to be confirmed):
openvinotoolkit/openvino_notebooks#2572
openvinotoolkit/openvino_notebooks#2619
openvinotoolkit/openvino_notebooks#2673
openvinotoolkit/openvino_notebooks#2663
openvinotoolkit/openvino_notebooks#2683
openvinotoolkit/openvino_notebooks#2686
openvinotoolkit/openvino_notebooks#2696

openvinotoolkit#2727 + openvinotoolkit#3211

nikita-savelyevv · 2025-01-30T14:53:40Z

ReleaseNotes.md

+  - Significantly faster data-free weight compression for OpenVINO models: INT4 compression is now up to 10x faster, while INT8 compression is up to 3x faster. The larger the model the higher the time reduction.
+  - AWQ weight compression is now up to 2x faster, improving overall runtime efficiency.
+  - Peak memory usage during INT4 data-free weight compression in the OpenVINO backend is reduced up to 50% for certain models.


#2727 + #3211

andreyanufr · 2025-01-31T07:27:07Z

no changes from my side

daniil-lyakhov · 2025-01-31T10:29:31Z

ReleaseNotes.md

+- General:
+  - ...
+- Features:
+  - (TorchFX, Experimental) Preview support for the new `quantize_pt2e` API has been introduced, enabling quantization of `torch.fx.GraphModule` models with the `OpenVINOQuantizer` and the `X86InductorQuantizer` quantizers. `quantize_pt2e` API utilizes `MinMax` algorithm statistic collectors, as well as `SmoothQuant`, `BiasCorrection` and `FastBiasCorrection` Post-Training Quantization algorithms.


#3121 + #3216

anzr299

No changes from me

andrey-churkin · 2025-01-31T18:37:27Z

ReleaseNotes.md

+  - ...
+- Features:
+  - (TorchFX, Experimental) Preview support for the new `quantize_pt2e` API has been introduced, enabling quantization of `torch.fx.GraphModule` models with the `OpenVINOQuantizer` and the `X86InductorQuantizer` quantizers. `quantize_pt2e` API utilizes `MinMax` algorithm statistic collectors, as well as `SmoothQuant`, `BiasCorrection` and `FastBiasCorrection` Post-Training Quantization algorithms.
+  - (TensorFlow) The `nncf.quantize()` method is now the recommended way for the quantization initialization for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.


andrey-churkin · 2025-01-31T18:39:20Z

ReleaseNotes.md

+  - AWQ weight compression is now up to 2x faster, improving overall runtime efficiency.
+  - Peak memory usage during INT4 data-free weight compression in the OpenVINO backend is reduced up to 50% for certain models.
+- Deprecations/Removals:
+  - (TensorFlow) The `nncf.tensorflow.create_compressed_model()` method is now marked as deprecated. Please use the `nncf.quantize()` method for the quantization initialization.


andrey-churkin · 2025-01-31T18:48:12Z

ReleaseNotes.md

+- Features:
+  - (TorchFX, Experimental) Preview support for the new `quantize_pt2e` API has been introduced, enabling quantization of `torch.fx.GraphModule` models with the `OpenVINOQuantizer` and the `X86InductorQuantizer` quantizers. `quantize_pt2e` API utilizes `MinMax` algorithm statistic collectors, as well as `SmoothQuant`, `BiasCorrection` and `FastBiasCorrection` Post-Training Quantization algorithms.
+  - (TensorFlow) The `nncf.quantize()` method is now the recommended way for the quantization initialization for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.
+  - (TensorFlow) Compression layers placement in the model now can be serialized and restored with new API functions: `nncf.tensorflow.get_config()` and `nncf.tensorflow.load_from_config()`. Please see [documentation](/docs/usage/training_time_compression/quantization_aware_training/Usage.md#saving-and-loading-compressed-models) for the saving/loading of a quantized model for more details.


alexsu52

LGTM

alexsu52 · 2025-02-03T07:36:45Z

ReleaseNotes.md

+  - ...
+- Features:
+  - (TorchFX, Experimental) Preview support for the new `quantize_pt2e` API has been introduced, enabling quantization of `torch.fx.GraphModule` models with the `OpenVINOQuantizer` and the `X86InductorQuantizer` quantizers. `quantize_pt2e` API utilizes `MinMax` algorithm statistic collectors, as well as `SmoothQuant`, `BiasCorrection` and `FastBiasCorrection` Post-Training Quantization algorithms.
+  - (TensorFlow) The `nncf.quantize()` method is now the recommended way for the quantization initialization for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.


Suggested change

- (TensorFlow) The `nncf.quantize()` method is now the recommended way for the quantization initialization for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.

- (TensorFlow) The `nncf.quantize()` method is now the recommended API for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.

nikita-malininn · 2025-02-03T10:10:57Z

@MaximProshin, release notes are ready for review.

### Changes - Added v2.15.0 template; ### Reason for changes - Upcoming release; ### Related tickets - 161230; #### For the contributors: Please add your changes (as the commit to the branch) to the list according to the template and previous notes; Do not add tests-related notes; Provide the list of the PRs (for all your notes) in the comment for the discussion; --------- Co-authored-by: Liubov Talamanova <[email protected]> Co-authored-by: Nikita Savelyev <[email protected]> Co-authored-by: Alexander Dokuchaev <[email protected]> Co-authored-by: Daniil Lyakhov <[email protected]> Co-authored-by: Andrey Churkin <[email protected]> (cherry picked from commit 80bd756)

(cherry picked from commit 80bd756)

nikita-malininn requested a review from a team as a code owner January 27, 2025 08:58

github-actions bot added documentation Improvements or additions to documentation release target labels Jan 27, 2025

nikita-malininn assigned alexsu52, ljaljushkin, l-bat, nikita-savelyevv, kshpv, daniil-lyakhov, andrey-churkin, andreyanufr, anzr299 and AlexanderDokuchaev Jan 27, 2025

Release notes template

df9b84d

ljaljushkin removed their assignment Jan 27, 2025

MaximProshin requested review from MaximProshin, alexsu52, l-bat, nikita-savelyevv, kshpv, AlexanderDokuchaev, anzr299, andrey-churkin, andreyanufr and daniil-lyakhov and removed request for a team January 28, 2025 08:21

Add list of OV notebooks with NNCF to release notes

11d2c8f

l-bat approved these changes Jan 28, 2025

View reviewed changes

Update ReleaseNotes.md

0d8826a

openvinotoolkit#2727 + openvinotoolkit#3211

nikita-savelyevv reviewed Jan 30, 2025

View reviewed changes

nikita-savelyevv approved these changes Jan 30, 2025

View reviewed changes

andreyanufr approved these changes Jan 31, 2025

View reviewed changes

Requirements

586a495

AlexanderDokuchaev approved these changes Jan 31, 2025

View reviewed changes

kshpv approved these changes Jan 31, 2025

View reviewed changes

AlexanderDokuchaev and others added 2 commits January 31, 2025 12:06

Update ReleaseNotes.md

4d98a64

quantize_pt2e and OpenVINOQuantizer

972acc9

daniil-lyakhov reviewed Jan 31, 2025

View reviewed changes

daniil-lyakhov approved these changes Jan 31, 2025

View reviewed changes

anzr299 approved these changes Jan 31, 2025

View reviewed changes

Update ReleaseNotes.md

75bf8a3

andrey-churkin reviewed Jan 31, 2025

View reviewed changes

Update ReleaseNotes.md

bc2fe54

andrey-churkin reviewed Jan 31, 2025

View reviewed changes

andrey-churkin self-requested a review January 31, 2025 18:48

andrey-churkin approved these changes Jan 31, 2025

View reviewed changes

alexsu52 approved these changes Feb 3, 2025

View reviewed changes

Update ReleaseNotes.md

07f5fb7

MaximProshin approved these changes Feb 3, 2025

View reviewed changes

nikita-malininn merged commit 80bd756 into openvinotoolkit:release_v2150 Feb 3, 2025
3 checks passed

Update ReleaseNotes.md

02c7732

nikita-malininn added a commit that referenced this pull request Feb 6, 2025

Update ReleaseNotes.md (#3214) (#3238)

879e80c

(cherry picked from commit 80bd756)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Release_v2150] Update ReleaseNotes.md #3214

[Release_v2150] Update ReleaseNotes.md #3214

nikita-malininn commented Jan 27, 2025

nikita-malininn commented Jan 27, 2025

ljaljushkin commented Jan 27, 2025

MaximProshin commented Jan 28, 2025

MaximProshin commented Jan 28, 2025

nikita-savelyevv Jan 30, 2025

andreyanufr commented Jan 31, 2025

daniil-lyakhov Jan 31, 2025

anzr299 left a comment •

edited

Loading

andrey-churkin Jan 31, 2025

andrey-churkin Jan 31, 2025

andrey-churkin Jan 31, 2025

andrey-churkin Jan 31, 2025

alexsu52 left a comment

alexsu52 Feb 3, 2025

nikita-malininn commented Feb 3, 2025

	- (TensorFlow) The `nncf.quantize()` method is now the recommended way for the quantization initialization for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.
	- (TensorFlow) The `nncf.quantize()` method is now the recommended API for Quantization-Aware Training. Please refer to an [example](examples/quantization_aware_training/tensorflow/mobilenet_v2) for more details about how to use new approach.

[Release_v2150] Update ReleaseNotes.md #3214

[Release_v2150] Update ReleaseNotes.md #3214

Conversation

nikita-malininn commented Jan 27, 2025

Changes

Reason for changes

Related tickets

For the contributors:

nikita-malininn commented Jan 27, 2025

ljaljushkin commented Jan 27, 2025

MaximProshin commented Jan 28, 2025

MaximProshin commented Jan 28, 2025

nikita-savelyevv Jan 30, 2025

Choose a reason for hiding this comment

andreyanufr commented Jan 31, 2025

daniil-lyakhov Jan 31, 2025

Choose a reason for hiding this comment

anzr299 left a comment • edited Loading

Choose a reason for hiding this comment

andrey-churkin Jan 31, 2025

Choose a reason for hiding this comment

andrey-churkin Jan 31, 2025

Choose a reason for hiding this comment

andrey-churkin Jan 31, 2025

Choose a reason for hiding this comment

andrey-churkin Jan 31, 2025

Choose a reason for hiding this comment

alexsu52 left a comment

Choose a reason for hiding this comment

alexsu52 Feb 3, 2025

Choose a reason for hiding this comment

nikita-malininn commented Feb 3, 2025

anzr299 left a comment •

edited

Loading