mct-nightly 2.4.0.20250630.629__tar.gz → 2.4.0.20250702.605__tar.gz

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mct-nightly
-Version: 2.4.0.20250630.629
+Version: 2.4.0.20250702.605
 Summary: A Model Compression Toolkit for neural networks
 Author-email: ssi-dnn-dev@sony.com
 Classifier: Programming Language :: Python :: 3
@@ -35,7 +35,7 @@ Dynamic: summary
 
 <div align="center" markdown="1">
 <p>
-      <a href="https://sony.github.io/model_optimization/" target="_blank">
+      <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization/" target="_blank">
         <img src="https://raw.githubusercontent.com/sony/model_optimization/refs/heads/main/docsrc/images/mctHeader1-cropped.svg" width="1000"></a>
 </p>
   
@@ -52,9 +52,9 @@ ______________________________________________________________________
   <a href="#license">License</a>
 </p>
 <p align="center">
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
   <a href="https://github.com/sony/model_optimization/releases"><img src="https://img.shields.io/github/v/release/sony/model_optimization" /></a>
   <a href="https://github.com/sony/model_optimization/blob/main/LICENSE.md"><img src="https://img.shields.io/badge/license-Apache%202.0-blue" /></a>
   
@@ -85,9 +85,9 @@ MCT supports various quantization methods as appears below.
 
   Quantization Method  | Complexity | Computational Cost | API | Tutorial 
 -------------------- | -----------|--------------------|---------|--------
-PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
-GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
-QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
+QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
 
 </p>    
 </div>
@@ -130,13 +130,13 @@ Generates synthetic images based on the statistics stored in the model's batch n
 The specifications of the method are detailed in the paper: _"**Data Generation for Hardware-Friendly Post-Training Quantization**"_ [5].
 __________________________________________________________________________________________________________
 ### Structured Pruning [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_pruning_mnist.ipynb)
-Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_pruning_experimental.html)).
+Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_pruning_experimental.html)).
 __________________________________________________________________________________________________________
 ### **Debugging and Visualization**
 **🎛️ Network Editor (Modify Quantization Configurations)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_network_editor.ipynb). 
 Modify your model's quantization configuration for specific layers or apply a custom edit rule (e.g adjust layer's bit-width) using MCT’s network editor.
 
-**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sony.github.io/model_optimization/guidelines/visualization.html).
+**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html).
 
 **🔑 XQuant (Explainable Quantization)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_xquant.ipynb). Get valuable insights regarding the quality and success of the quantization process of your model. The report includes histograms and similarity metrics between the original float model and the quantized model in key points of the model. The report can be visualized using TensorBoard.
 __________________________________________________________________________________________________________
@@ -146,15 +146,15 @@ The specifications of the algorithm are detailed in the paper: _"**EPTQ: Enhance
 More details on how to use EPTQ via MCT can be found in the [GPTQ guidelines](https://github.com/sony/model_optimization/blob/main/model_compression_toolkit/gptq/README.md).
 
 ## <div align="center">Resources</div>
-* [User Guide](https://sony.github.io/model_optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
+* [User Guide](https://sonysemiconductorsolutions.github.io/mct-model-optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
 
-* MCT's [API Docs](https://sony.github.io/model_optimization/api/api_docs/) is separated per quantization methods:  
+* MCT's [API Docs](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/) is separated per quantization methods:  
 
-  * [Post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#ptq) | PTQ API docs
-  * [Gradient-based post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#gptq) | GPTQ API docs
-  * [Quantization-aware training](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | QAT API docs
+  * [Post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#ptq) | PTQ API docs
+  * [Gradient-based post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#gptq) | GPTQ API docs
+  * [Quantization-aware training](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | QAT API docs
     
-* [Debug](https://sony.github.io/model_optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
+* [Debug](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
   
 * [Release notes](https://github.com/sony/model_optimization/releases)
 

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/README.md

@@ -1,6 +1,6 @@
 <div align="center" markdown="1">
 <p>
-      <a href="https://sony.github.io/model_optimization/" target="_blank">
+      <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization/" target="_blank">
         <img src="https://raw.githubusercontent.com/sony/model_optimization/refs/heads/main/docsrc/images/mctHeader1-cropped.svg" width="1000"></a>
 </p>
   
@@ -17,9 +17,9 @@ ______________________________________________________________________
   <a href="#license">License</a>
 </p>
 <p align="center">
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
   <a href="https://github.com/sony/model_optimization/releases"><img src="https://img.shields.io/github/v/release/sony/model_optimization" /></a>
   <a href="https://github.com/sony/model_optimization/blob/main/LICENSE.md"><img src="https://img.shields.io/badge/license-Apache%202.0-blue" /></a>
   
@@ -50,9 +50,9 @@ MCT supports various quantization methods as appears below.
 
   Quantization Method  | Complexity | Computational Cost | API | Tutorial 
 -------------------- | -----------|--------------------|---------|--------
-PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
-GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
-QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
+QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
 
 </p>    
 </div>
@@ -95,13 +95,13 @@ Generates synthetic images based on the statistics stored in the model's batch n
 The specifications of the method are detailed in the paper: _"**Data Generation for Hardware-Friendly Post-Training Quantization**"_ [5].
 __________________________________________________________________________________________________________
 ### Structured Pruning [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_pruning_mnist.ipynb)
-Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_pruning_experimental.html)).
+Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_pruning_experimental.html)).
 __________________________________________________________________________________________________________
 ### **Debugging and Visualization**
 **🎛️ Network Editor (Modify Quantization Configurations)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_network_editor.ipynb). 
 Modify your model's quantization configuration for specific layers or apply a custom edit rule (e.g adjust layer's bit-width) using MCT’s network editor.
 
-**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sony.github.io/model_optimization/guidelines/visualization.html).
+**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html).
 
 **🔑 XQuant (Explainable Quantization)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_xquant.ipynb). Get valuable insights regarding the quality and success of the quantization process of your model. The report includes histograms and similarity metrics between the original float model and the quantized model in key points of the model. The report can be visualized using TensorBoard.
 __________________________________________________________________________________________________________
@@ -111,15 +111,15 @@ The specifications of the algorithm are detailed in the paper: _"**EPTQ: Enhance
 More details on how to use EPTQ via MCT can be found in the [GPTQ guidelines](https://github.com/sony/model_optimization/blob/main/model_compression_toolkit/gptq/README.md).
 
 ## <div align="center">Resources</div>
-* [User Guide](https://sony.github.io/model_optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
+* [User Guide](https://sonysemiconductorsolutions.github.io/mct-model-optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
 
-* MCT's [API Docs](https://sony.github.io/model_optimization/api/api_docs/) is separated per quantization methods:  
+* MCT's [API Docs](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/) is separated per quantization methods:  
 
-  * [Post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#ptq) | PTQ API docs
-  * [Gradient-based post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#gptq) | GPTQ API docs
-  * [Quantization-aware training](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | QAT API docs
+  * [Post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#ptq) | PTQ API docs
+  * [Gradient-based post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#gptq) | GPTQ API docs
+  * [Quantization-aware training](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | QAT API docs
     
-* [Debug](https://sony.github.io/model_optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
+* [Debug](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
   
 * [Release notes](https://github.com/sony/model_optimization/releases)
 

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/mct_nightly.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mct-nightly
-Version: 2.4.0.20250630.629
+Version: 2.4.0.20250702.605
 Summary: A Model Compression Toolkit for neural networks
 Author-email: ssi-dnn-dev@sony.com
 Classifier: Programming Language :: Python :: 3
@@ -35,7 +35,7 @@ Dynamic: summary
 
 <div align="center" markdown="1">
 <p>
-      <a href="https://sony.github.io/model_optimization/" target="_blank">
+      <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization/" target="_blank">
         <img src="https://raw.githubusercontent.com/sony/model_optimization/refs/heads/main/docsrc/images/mctHeader1-cropped.svg" width="1000"></a>
 </p>
   
@@ -52,9 +52,9 @@ ______________________________________________________________________
   <a href="#license">License</a>
 </p>
 <p align="center">
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
-  <a href="https://sony.github.io/model_optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/pytorch-2.3%20%7C%202.4%20%7C%202.5%20%7C%202.6-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/tensorflow-2.14%20%7C%202.15-blue" /></a>
+  <a href="https://sonysemiconductorsolutions.github.io/mct-model-optimization#prerequisites"><img src="https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue" /></a>
   <a href="https://github.com/sony/model_optimization/releases"><img src="https://img.shields.io/github/v/release/sony/model_optimization" /></a>
   <a href="https://github.com/sony/model_optimization/blob/main/LICENSE.md"><img src="https://img.shields.io/badge/license-Apache%202.0-blue" /></a>
   
@@ -85,9 +85,9 @@ MCT supports various quantization methods as appears below.
 
   Quantization Method  | Complexity | Computational Cost | API | Tutorial 
 -------------------- | -----------|--------------------|---------|--------
-PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
-GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
-QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+PTQ (Post Training Quantization)  | Low | Low (~1-10 CPU minutes) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_post_training_quantization.ipynb"><img src="https://img.shields.io/badge/Pytorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_post-training_quantization.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
+GPTQ (parameters fine-tuning using gradients)  | Moderate | Moderate (~1-3 GPU hours) | [PyTorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_gradient_post_training_quantization.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_gradient_post_training_quantization.html) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/PyTorch-green"/></a> <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_mobilenet_gptq.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a> 
+QAT (Quantization Aware Training)  | High | High (~12-36 GPU hours) | [QAT API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | <a href="https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_qat.ipynb"><img src="https://img.shields.io/badge/Keras-green"/></a>
 
 </p>    
 </div>
@@ -130,13 +130,13 @@ Generates synthetic images based on the statistics stored in the model's batch n
 The specifications of the method are detailed in the paper: _"**Data Generation for Hardware-Friendly Post-Training Quantization**"_ [5].
 __________________________________________________________________________________________________________
 ### Structured Pruning [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_pruning_mnist.ipynb)
-Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sony.github.io/model_optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sony.github.io/model_optimization/api/api_docs/methods/keras_pruning_experimental.html)).
+Reduces model size/complexity and ensures better channels utilization by removing redundant input channels from layers and reconstruction of layer weights. Read more ([Pytorch API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/pytorch_pruning_experimental.html) / [Keras API](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/methods/keras_pruning_experimental.html)).
 __________________________________________________________________________________________________________
 ### **Debugging and Visualization**
 **🎛️ Network Editor (Modify Quantization Configurations)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/keras/example_keras_network_editor.ipynb). 
 Modify your model's quantization configuration for specific layers or apply a custom edit rule (e.g adjust layer's bit-width) using MCT’s network editor.
 
-**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sony.github.io/model_optimization/guidelines/visualization.html).
+**🖥️ Visualization**. Observe useful information for troubleshooting the quantized model's performance using TensorBoard. [Read more](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html).
 
 **🔑 XQuant (Explainable Quantization)** [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sony/model_optimization/blob/main/tutorials/notebooks/mct_features_notebooks/pytorch/example_pytorch_xquant.ipynb). Get valuable insights regarding the quality and success of the quantization process of your model. The report includes histograms and similarity metrics between the original float model and the quantized model in key points of the model. The report can be visualized using TensorBoard.
 __________________________________________________________________________________________________________
@@ -146,15 +146,15 @@ The specifications of the algorithm are detailed in the paper: _"**EPTQ: Enhance
 More details on how to use EPTQ via MCT can be found in the [GPTQ guidelines](https://github.com/sony/model_optimization/blob/main/model_compression_toolkit/gptq/README.md).
 
 ## <div align="center">Resources</div>
-* [User Guide](https://sony.github.io/model_optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
+* [User Guide](https://sonysemiconductorsolutions.github.io/mct-model-optimization/index.html)  contains detailed information about MCT and guides you from installation through optimizing models for your edge AI applications.
 
-* MCT's [API Docs](https://sony.github.io/model_optimization/api/api_docs/) is separated per quantization methods:  
+* MCT's [API Docs](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/) is separated per quantization methods:  
 
-  * [Post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#ptq) | PTQ API docs
-  * [Gradient-based post-training quantization](https://sony.github.io/model_optimization/api/api_docs/index.html#gptq) | GPTQ API docs
-  * [Quantization-aware training](https://sony.github.io/model_optimization/api/api_docs/index.html#qat) | QAT API docs
+  * [Post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#ptq) | PTQ API docs
+  * [Gradient-based post-training quantization](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#gptq) | GPTQ API docs
+  * [Quantization-aware training](https://sonysemiconductorsolutions.github.io/mct-model-optimization/api/api_docs/index.html#qat) | QAT API docs
     
-* [Debug](https://sony.github.io/model_optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
+* [Debug](https://sonysemiconductorsolutions.github.io/mct-model-optimization/guidelines/visualization.html) – modify optimization process or generate an explainable report
   
 * [Release notes](https://github.com/sony/model_optimization/releases)
 

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/mct_nightly.egg-info/SOURCES.txt

@@ -118,7 +118,6 @@ model_compression_toolkit/core/common/quantization/filter_nodes_candidates.py
 model_compression_toolkit/core/common/quantization/node_quantization_config.py
 model_compression_toolkit/core/common/quantization/quantization_config.py
 model_compression_toolkit/core/common/quantization/quantization_fn_selection.py
-model_compression_toolkit/core/common/quantization/quantization_params_fn_selection.py
 model_compression_toolkit/core/common/quantization/quantize_graph_weights.py
 model_compression_toolkit/core/common/quantization/quantize_node.py
 model_compression_toolkit/core/common/quantization/set_node_quantization_config.py
@@ -211,9 +210,10 @@ model_compression_toolkit/core/keras/mixed_precision/configurable_activation_qua
 model_compression_toolkit/core/keras/mixed_precision/configurable_weights_quantizer.py
 model_compression_toolkit/core/keras/pruning/__init__.py
 model_compression_toolkit/core/keras/pruning/pruning_keras_implementation.py
-model_compression_toolkit/core/keras/quantizer/__init__.py
-model_compression_toolkit/core/keras/quantizer/fake_quant_builder.py
-model_compression_toolkit/core/keras/quantizer/lut_fake_quant.py
+model_compression_toolkit/core/keras/quantization/__init__.py
+model_compression_toolkit/core/keras/quantization/activation_quantization_fn_factory.py
+model_compression_toolkit/core/keras/quantization/fake_quant_builder.py
+model_compression_toolkit/core/keras/quantization/lut_fake_quant.py
 model_compression_toolkit/core/keras/reader/__init__.py
 model_compression_toolkit/core/keras/reader/common.py
 model_compression_toolkit/core/keras/reader/connectivity_handler.py
@@ -281,9 +281,10 @@ model_compression_toolkit/core/pytorch/mixed_precision/configurable_activation_q
 model_compression_toolkit/core/pytorch/mixed_precision/configurable_weights_quantizer.py
 model_compression_toolkit/core/pytorch/pruning/__init__.py
 model_compression_toolkit/core/pytorch/pruning/pruning_pytorch_implementation.py
-model_compression_toolkit/core/pytorch/quantizer/__init__.py
-model_compression_toolkit/core/pytorch/quantizer/fake_quant_builder.py
-model_compression_toolkit/core/pytorch/quantizer/lut_fake_quant.py
+model_compression_toolkit/core/pytorch/quantization/__init__.py
+model_compression_toolkit/core/pytorch/quantization/activation_quantization_fn_factory.py
+model_compression_toolkit/core/pytorch/quantization/fake_quant_builder.py
+model_compression_toolkit/core/pytorch/quantization/lut_fake_quant.py
 model_compression_toolkit/core/pytorch/reader/__init__.py
 model_compression_toolkit/core/pytorch/reader/graph_builders.py
 model_compression_toolkit/core/pytorch/reader/node_holders.py
@@ -439,6 +440,8 @@ model_compression_toolkit/qat/pytorch/quantizer/lsq/uniform_lsq.py
 model_compression_toolkit/qat/pytorch/quantizer/ste_rounding/__init__.py
 model_compression_toolkit/qat/pytorch/quantizer/ste_rounding/symmetric_ste.py
 model_compression_toolkit/qat/pytorch/quantizer/ste_rounding/uniform_ste.py
+model_compression_toolkit/quantization_preparation/__init__.py
+model_compression_toolkit/quantization_preparation/load_fqc.py
 model_compression_toolkit/target_platform_capabilities/__init__.py
 model_compression_toolkit/target_platform_capabilities/constants.py
 model_compression_toolkit/target_platform_capabilities/immutable.py

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/model_compression_toolkit/__init__.py

@@ -27,4 +27,4 @@ from model_compression_toolkit import data_generation
 from model_compression_toolkit import pruning
 from model_compression_toolkit.trainable_infrastructure.keras.load_model import keras_load_quantized_model
 
-__version__ = "2.4.0.20250630.000629"
+__version__ = "2.4.0.20250702.000605"

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/model_compression_toolkit/core/common/back2framework/base_model_builder.py

@@ -15,7 +15,6 @@
 from abc import ABC, abstractmethod
 from typing import Any, Tuple
 
-from model_compression_toolkit.core.common.framework_info import FrameworkInfo
 from model_compression_toolkit.core import common
 from model_compression_toolkit.core.common.user_info import UserInformation
 

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/model_compression_toolkit/core/common/framework_info.py

@@ -13,20 +13,10 @@
 # limitations under the License.
 # ==============================================================================
 
-
-from collections.abc import Callable
 from enum import Enum
-from typing import Dict, Any, Tuple, NamedTuple
+from typing import Dict, Any, Tuple, NamedTuple, Optional
 from abc import ABC, abstractmethod
 
-from mct_quantizers import QuantizationMethod
-
-
-# Default value to use for ops without kernel.
-# This is a weird default, but it's used all over the place, so for now only extract it to const so that it can be
-# referenced by variable instead of hard-coded.
-DEFAULT_KERNEL_ATTRIBUTE = None
-
 
 class ChannelAxis(Enum):
     """
@@ -58,24 +48,21 @@ class FrameworkInfo(ABC):
     no_quantization_ops:Layers that should not get quantized (e.g., Reshape, Transpose, etc.)
 
     Fields:
-        activation_quantizer_mapping (Dict[QuantizationMethod, Callable]): A dictionary mapping from QuantizationMethod to a quantization function.
         kernel_channels_mapping (Dict): Dictionary from a layer to a tuple of its kernel in/out channels indices.
         kernel_ops_attribute_mapping (Dict): Dictionary from a framework operator to its weight attribute to quantize.
         out_channel_axis_mapping (Dict): Dictionary of output channels of the model's layers (for computing statistics per-channel).
         _layer_min_max_mapping (Dict[Any, tuple]): Dictionary from a layer to its min/max output values.
-
     """
 
-    activation_quantizer_mapping: Dict[QuantizationMethod, Callable]
-    kernel_channels_mapping: Dict[Any, ChannelAxisMapping]
     kernel_ops_attribute_mapping: Dict[Any, str]
+    kernel_channels_mapping: Dict[Any, ChannelAxisMapping]
     out_channel_axis_mapping: Dict[Any, int]
-    _layer_min_max_mapping: Dict[Any, tuple]
 
+    _layer_min_max_mapping: Dict[Any, tuple]
     _default_channel_mapping = ChannelAxisMapping(None, None)
 
     @classmethod
-    def get_kernel_op_attribute(cls, node_type: Any) -> str:
+    def get_kernel_op_attribute(cls, node_type: Any) -> Optional[str]:
         """
         Get attribute of a layer's weight to quantize.
 
@@ -85,20 +72,7 @@ class FrameworkInfo(ABC):
         Returns:
             Attribute the layer has and should be quantized.
         """
-        return cls.kernel_ops_attribute_mapping.get(node_type, DEFAULT_KERNEL_ATTRIBUTE)
-
-    @classmethod
-    def is_kernel_op(cls, node_type: Any) -> bool:
-        """
-        Check is the node is a kernel operation.
-
-        Args:
-            node_type: Layer to get its attributes.
-
-        Returns:
-            True if node type is a kernel operation, else False.
-        """
-        return node_type in cls.kernel_ops_attribute_mapping
+        return cls.kernel_ops_attribute_mapping.get(node_type)
 
     @classmethod
     def get_layer_min_max(cls, layer: Any, fw_attrs: Dict) -> Tuple[float, float]:
@@ -169,7 +143,6 @@ def get_fw_info():
     Returns: FrameworkInfo class.
     """
     assert _current_framework_info is not None, "fw_info isn't initialized."
-    assert issubclass(_current_framework_info, FrameworkInfo), "fw_info isn't initialized to a FrameworkInfo class."
     return _current_framework_info
 
 

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/model_compression_toolkit/core/common/fusion/graph_fuser.py

@@ -14,12 +14,12 @@
 #  ==============================================================================
 
 import copy
-from typing import List, Tuple
+from typing import Tuple
 
 from model_compression_toolkit.core.common.fusion.fusing_info import FusingInfoGenerator
 from model_compression_toolkit.core.common.graph.base_graph import Graph, BaseNode, OutTensor
-from model_compression_toolkit.core.common.quantization.candidate_node_quantization_config import CandidateNodeQuantizationConfig
-from itertools import product
+from model_compression_toolkit.core.common.quantization.candidate_node_quantization_config import \
+    CandidateNodeQuantizationConfig, NodeQuantizationConfig
 
 
 class FusedLayerType:
@@ -30,6 +30,7 @@ class FusedLayerType:
     def __init__(self):
         self.__name__ = 'FusedLayer'
 
+
 class GraphFuser:
     def apply_node_fusion(self, graph: Graph) -> Graph:
         """
@@ -64,7 +65,6 @@ class GraphFuser:
 
         return graph_copy
 
-
     @staticmethod
     def _create_fused_node(fused_node_id: str, nodes: Tuple[BaseNode]) -> BaseNode:
         """
@@ -86,10 +86,15 @@ class GraphFuser:
                               weights={},
                               layer_class=FusedLayerType)
 
+        base_cfg = CandidateNodeQuantizationConfig(
+            activation_quantization_cfg=nodes[-1].quantization_cfg.base_quantization_cfg.activation_quantization_cfg,
+            weights_quantization_cfg=None
+        )
         activation_cfgs = [c.activation_quantization_cfg for c in nodes[-1].candidates_quantization_cfg]
-        fused_node.candidates_quantization_cfg = [
-            CandidateNodeQuantizationConfig(weights_quantization_cfg=None, activation_quantization_cfg=a) for a in
-            activation_cfgs]
+        candidates = [CandidateNodeQuantizationConfig(weights_quantization_cfg=None, activation_quantization_cfg=a)
+                      for a in activation_cfgs]
+        fused_node.quantization_cfg = NodeQuantizationConfig(base_quantization_cfg=base_cfg,
+                                                             candidates_quantization_cfg=candidates)
 
         # Keep the final configurations if they were set already.
         fused_node.final_weights_quantization_cfg = nodes[0].final_weights_quantization_cfg
@@ -158,5 +163,3 @@ class GraphFuser:
 
         # Finally, add the new fused node to the graph
         graph.add_node(fused_node)
-
-

{mct_nightly-2.4.0.20250630.629 → mct_nightly-2.4.0.20250702.605}/model_compression_toolkit/core/common/graph/base_graph.py

@@ -39,6 +39,7 @@ from model_compression_toolkit.target_platform_capabilities.targetplatform2frame
 from model_compression_toolkit.target_platform_capabilities.targetplatform2framework.framework_quantization_capabilities import \
     FrameworkQuantizationCapabilities
 
+
 def validate_graph_after_change(method: Callable) -> Callable:
     """
     Decorator for graph-mutating methods. After the decorated method executes,
@@ -120,28 +121,13 @@ class Graph(nx.MultiDiGraph, GraphSearches):
     def fusing_info(self, fusing_info: FusingInfo):
         self._fusing_info = fusing_info
 
-    def set_fqc(self,
-                fqc: FrameworkQuantizationCapabilities):
+    def set_fqc(self, fqc: FrameworkQuantizationCapabilities):
         """
         Set the graph's FQC.
         Args:
             fqc: FrameworkQuantizationCapabilities object.
         """
-        # validate graph nodes are either from the framework or a custom layer defined in the FQC
-        # Validate graph nodes are either built-in layers from the framework or custom layers defined in the FQC
-        fqc_layers = fqc.op_sets_to_layers.get_layers()
-        fqc_filtered_layers = [layer for layer in fqc_layers if isinstance(layer, LayerFilterParams)]
-        for n in self.nodes:
-            is_node_in_fqc = any([n.is_match_type(_type) for _type in fqc_layers]) or \
-                             any([n.is_match_filter_params(filtered_layer) for filtered_layer in fqc_filtered_layers])
-            if n.is_custom:
-                if not is_node_in_fqc:
-                    Logger.critical(f'MCT does not support optimizing Keras custom layers. Found a layer of type {n.type}. '
-                                    ' Please add the custom layer to Framework Quantization Capabilities (FQC), or file a feature '
-                                    'request or an issue if you believe this should be supported.')  # pragma: no cover
-                if any([qc.default_weight_attr_config.enable_weights_quantization for qc in n.get_qco(fqc).quantization_configurations]):
-                    Logger.critical(f'Layer identified: {n.type}. MCT does not support weight quantization for Keras custom layers.')  # pragma: no cover
-
+        # TODO irena: this is only passed for negative shift activation.
         self.fqc = fqc
 
     def get_topo_sorted_nodes(self):
@@ -578,7 +564,7 @@ class Graph(nx.MultiDiGraph, GraphSearches):
             A list of nodes that their weights can be configured (namely, has one or more weight qc candidate).
         """
         # configurability is only relevant for kernel attribute quantization
-        potential_conf_nodes = [n for n in list(self) if n.is_kernel_op]
+        potential_conf_nodes = [n for n in self.nodes if n.kernel_attr]
 
         def is_configurable(n):
             return n.is_configurable_weight(n.kernel_attr) and (not n.reuse or include_reused_nodes)
@@ -693,10 +679,8 @@ class Graph(nx.MultiDiGraph, GraphSearches):
         """
         Gets the final number of bits for quantization of each weights' configurable layer.
 
-        Args:
-            fw_info: fw_info: FrameworkInfo object with information about the specific framework's model.
-
-        Returns: A list of pairs of (node type, node's weights quantization bitwidth).
+        Returns:
+            A list of pairs of (node type, node's weights quantization bitwidth).
 
         """
         sorted_conf_weights = self.get_sorted_weights_configurable_nodes()
@@ -876,32 +860,31 @@ class Graph(nx.MultiDiGraph, GraphSearches):
 
         return intermediate_nodes, next_node
 
+    # TODO irena move to load_fqc and clean up tests (currently tests_pytest/common_tests/unit_tests/core/graph/test_base_graph.py)
     def override_fused_node_activation_quantization_candidates(self):
         """
         Override fused node activation quantization candidates for all nodes in fused operations,
         except for the last node in each fused group.
         Update the value of quantization_config with the value of op_quaitization_cfg from FusingInfo.
         """
-        from model_compression_toolkit.core.common.quantization.candidate_node_quantization_config import CandidateNodeQuantizationConfig
-
         nodes_in_fln = self.fusing_info.get_inner_fln_nodes()
         for node in nodes_in_fln:
             fused_node_op_id = self.fusing_info.get_fused_op_id_for_node(node.name)
-            fusiong_op_quaitization_cfg = self.fusing_info.get_fused_op_quantization_config(fused_node_op_id)             
-            org_candidate = node.candidates_quantization_cfg[0]
-            if fusiong_op_quaitization_cfg is not None and fusiong_op_quaitization_cfg.enable_activation_quantization:
-                # Set ActivationQuantizationMode to FLN_QUANT and update the value of quantization_config
-                activation_quantization_cfg = NodeActivationQuantizationConfig(qc=org_candidate,
-                                                                               op_cfg=fusiong_op_quaitization_cfg,
-                                                                               activation_quantization_fn=org_candidate.activation_quantization_cfg.activation_quantization_fn,
-                                                                               activation_quantization_params_fn=org_candidate.activation_quantization_cfg.activation_quantization_params_fn)
-                activation_quantization_cfg.quant_mode = ActivationQuantizationMode.FLN_QUANT
-                for qc in node.candidates_quantization_cfg:
-                    qc.activation_quantization_cfg = activation_quantization_cfg
+            fusing_op_quantization_cfg = self.fusing_info.get_fused_op_quantization_config(fused_node_op_id)
+            if fusing_op_quantization_cfg is not None and fusing_op_quantization_cfg.enable_activation_quantization:
+                def update(qc):
+                    qc.activation_quantization_cfg = NodeActivationQuantizationConfig(fusing_op_quantization_cfg)
+                    qc.activation_quantization_cfg.quant_mode = ActivationQuantizationMode.FLN_QUANT
+                node.quantization_cfg.update_all(update, remove_duplicates=True)
             else:
-                # Set ActivationQuantizationMode to FLN_NO_QUANT
+                node.quantization_cfg.update_activation_quantization_mode(ActivationQuantizationMode.FLN_NO_QUANT)
+                # Remove duplicate candidates. We cannot compare whole candidates since activation configs might not
+                # be identical, but we do want to treat them as such. So we only check duplication by weight configs.
+                uniq_qcs = []
                 for qc in node.candidates_quantization_cfg:
-                    qc.activation_quantization_cfg.quant_mode = ActivationQuantizationMode.FLN_NO_QUANT
+                    if not any(qc.weights_quantization_cfg == uqc.weights_quantization_cfg for uqc in uniq_qcs):
+                        uniq_qcs.append(qc)
+                node.quantization_cfg.candidates_quantization_cfg = uniq_qcs
 
     def validate(self):
         """