Release v2.9.0 of NNCF to master

.github/workflows/mypy.yml

@@ -0,0 +1,23 @@
+name: mypy
+
+on:
+  pull_request:
+    types:
+      - opened
+      - reopened
+      - synchronize
+
+jobs:
+  mypy:
+    runs-on: ubuntu-20.04
+    steps:
+    - uses: actions/checkout@v3
+    - uses: actions/setup-python@v3
+      with:
+        python-version: 3.8.10
+    - name: Install NNCF
+      run: pip install -e .
+    - name: Install mypy
+      run: pip install mypy==1.8.0
+    - name: Run mypy
+      run: mypy --install-types --config-file=.mypy.ini --non-interactive

.mypy.ini

@@ -0,0 +1,9 @@
+[mypy]
+files = nncf/common/sparsity
+follow_imports = silent
+strict = True
+
+# should be removed later
+# mypy recommends the following tool as an autofix:
+# https://github.com/hauntsaninja/no_implicit_optional
+implicit_optional = True

.pre-commit-config.yaml

@@ -26,3 +26,4 @@ repos:
     hooks:
       - id: markdownlint
         args: [--config=.markdownlint.yaml]
+

Makefile

@@ -23,6 +23,8 @@ install-pre-commit:
 install-onnx-test:
 	pip install -U pip
 	pip install -e .[onnx]
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=omz-tools&subdirectory=tools/model_tools"
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=accuracy_checker&subdirectory=tools/accuracy_checker"
 	pip install -r tests/onnx/requirements.txt
 	pip install -r tests/cross_fw/install/requirements.txt
 	pip install -r tests/cross_fw/examples/requirements.txt
@@ -50,14 +52,14 @@ test-examples-onnx:
 install-openvino-test:
 	pip install -U pip
 	pip install -e .[openvino]
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=omz-tools&subdirectory=tools/model_tools"
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=accuracy_checker&subdirectory=tools/accuracy_checker"
 	pip install tensorflow==2.12.0
 	pip install -r tests/openvino/requirements.txt
 	pip install -r tests/cross_fw/install/requirements.txt
 	pip install -r tests/cross_fw/examples/requirements.txt
 
 install-openvino-dev: install-openvino-test install-pre-commit
-	pip install -r examples/experimental/openvino/bert/requirements.txt
-	pip install -r examples/experimental/openvino/yolo_v5/requirements.txt
 	pip install -r examples/post_training_quantization/openvino/mobilenet_v2/requirements.txt
 	pip install -r examples/post_training_quantization/openvino/anomaly_stfpm_quantize_with_accuracy_control/requirements.txt
 	pip install -r examples/post_training_quantization/openvino/yolov8/requirements.txt
@@ -81,6 +83,8 @@ test-examples-openvino:
 install-tensorflow-test:
 	pip install -U pip
 	pip install -e .[tf]
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=omz-tools&subdirectory=tools/model_tools"
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=accuracy_checker&subdirectory=tools/accuracy_checker"
 	pip install -r tests/tensorflow/requirements.txt
 	pip install -r tests/cross_fw/install/requirements.txt
 	pip install -r tests/cross_fw/examples/requirements.txt
@@ -105,6 +109,8 @@ test-examples-tensorflow:
 install-torch-test:
 	pip install -U pip
 	pip install -e .[torch] --index-url https://download.pytorch.org/whl/cu118 --extra-index-url=https://pypi.org/simple  # ticket 119128
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=omz-tools&subdirectory=tools/model_tools"
+	pip install "git+https://github.com/openvinotoolkit/open_model_zoo.git@37f60eb#egg=accuracy_checker&subdirectory=tools/accuracy_checker"
 	pip install -r tests/torch/requirements.txt --index-url https://download.pytorch.org/whl/cu118 --extra-index-url=https://pypi.org/simple
 	pip install -r tests/cross_fw/install/requirements.txt
 	pip install -r tests/cross_fw/examples/requirements.txt
@@ -170,3 +176,12 @@ test-examples:
 # Pre commit check
 pre-commit:
 	pre-commit run -a
+
+
+###############################################################################
+# Fuzzing tests
+install-fuzz-test: install-common-test
+	pip install -r tests/cross_fw/sdl/fuzz/requirements.txt
+
+test-fuzz:
+	python tests/cross_fw/sdl/fuzz/quantize_api.py

ReleaseNotes.md

@@ -1,5 +1,42 @@
 # Release Notes
 
+## New in Release 2.9.0
+
+Post-training Quantization:
+
+- Features:
+  - (OpenVINO) Added modified AWQ algorithm for 4-bit data-aware weights compression. This algorithm applied only for patterns `MatMul->Multiply->Matmul`. For that `awq` optional parameter has been added to `nncf.compress_weights()` and can be used to minimize accuracy degradation of compressed models (note that this option increases the compression time).
+  - (ONNX) Introduced support for the ONNX backend in the `nncf.quantize_with_accuracy_control()` method. Users can now perform quantization with accuracy control for `onnx.ModelProto`. By leveraging this feature, users can enhance the accuracy of quantized models while minimizing performance impact.
+  - (ONNX) Added an example based on the YOLOv8n-seg model for demonstrating the usage of quantization with accuracy control for the ONNX backend.
+  - (PT) Added SmoothQuant algorithm for PyTorch backend in `nncf.quantize()`.
+  - (OpenVINO) Added [an example](examples/llm_compression/openvino/tiny_llama_find_hyperparams) with the hyperparameters tuning for the TinyLLama model.
+  - Introduced the `nncf.AdvancedAccuracyRestorerParameters`.
+  - Introduced the `subset_size` option for the `nncf.compress_weights()`.
+  - Introduced `TargetDevice.NPU` as the replacement for `TargetDevice.VPU`.
+- Fixes:
+  - Fixed API Enums serialization/deserialization issue.
+  - Fixed issue with required arguments for `revert_operations_to_floating_point_precision` method.
+- Improvements:
+  - (ONNX) Aligned statistics collection with OpenVINO and PyTorch backends.
+  - Extended `nncf.compress_weights()` with Convolution & Embeddings compression in order to reduce memory footprint.
+- Deprecations/Removals:
+  - (OpenVINO) Removed outdated examples with `nncf.quantize()` for BERT and YOLOv5 models.
+  - (OpenVINO) Removed outdated example with `nncf.quantize_with_accuracy_control()` for SSD MobileNetV1 FPN model.
+  - (PyTorch) Deprecated the `binarization` algorithm.
+  - Removed Post-training Optimization Tool as OpenVINO backend.
+  - Removed Dockerfiles.
+  - `TargetDevice.VPU` was replaced by `TargetDevice.NPU`.
+- Tutorials:
+  - [Post-Training Optimization of Stable Diffusion v2 Model](https://github.com/openvinotoolkit/openvino_notebooks/tree/main/notebooks/236-stable-diffusion-v2/236-stable-diffusion-v2-text-to-image.ipynb)
+  - [Post-Training Optimization of DeciDiffusion Model](https://github.com/openvinotoolkit/openvino_notebooks/tree/main/notebooks/259-decidiffusion-image-generation/259-decidiffusion-image-generation.ipynb)
+  - [Post-Training Optimization of DepthAnything Model](https://github.com/openvinotoolkit/openvino_notebooks/tree/main/notebooks/280-depth-anything/280-depth-anything.ipynb)
+  - [Post-Training Optimization of Stable Diffusion ControlNet Model](https://github.com/openvinotoolkit/openvino_notebooks/tree/main/notebooks/235-controlnet-stable-diffusion/235-controlnet-stable-diffusion.ipynb)
+
+Compression-aware training:
+
+- Fixes
+  - (PyTorch) Fixed issue with `NNCFNetworkInterface.get_clean_shallow_copy` missed arguments.
+
 ## New in Release 2.8.1
 
 Post-training Quantization:
@@ -22,7 +59,7 @@ Post-training Quantization:
   - (OpenVINO) Changed default OpenVINO opset from 9 to 13.
 - Features:
   - (OpenVINO) Added 4-bit data-aware weights compression. For that `dataset` optional parameter has been added to `nncf.compress_weights()` and can be used to minimize accuracy degradation of compressed models (note that this option increases the compression time).
-  - (PyTorch) Added support for PyTorch models with shared weights and custom PyTorch modules in `nncf.compress_weights()`. The weights compression algorithm for PyTorch models is now based on tracing the model graph. The `dataset` parameter is now required in `nncf.compress_weights()` for the compression of PyTorch models.
+  - (PyTorch) Added support for PyTorch models with shared weights and custom PyTorch modules in nncf.compress_weights(). The weights compression algorithm for PyTorch models is now based on tracing the model graph. The dataset parameter is now required in nncf.compress_weights() for the compression of PyTorch models.
   - (Common) Renamed the `nncf.CompressWeightsMode.INT8` to `nncf.CompressWeightsMode.INT8_ASYM` and introduce `nncf.CompressWeightsMode.INT8_SYM` that can be efficiently used with dynamic 8-bit quantization of activations.
   The original `nncf.CompressWeightsMode.INT8` enum value is now deprecated.
   - (OpenVINO) Added support for quantizing the ScaledDotProductAttention operation from OpenVINO opset 13.