Remove MIXED quantization preset for transformer models (#1728)

Since recently MIXED preset is set by default for transformer models.

List of notebooks where preset is set manually:

[122-yolov8-quantization-with-accuracy-control.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/122-quantizing-model-with-accuracy-control/122-yolov8-quantization-with-accuracy-control.ipynb)
– not a transformer

[226-yolov7-optimization.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/226-yolov7-optimization/226-yolov7-optimization.ipynb)
– not a transformer

[230-yolov8-instance-segmentation.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/230-yolov8-optimization/230-yolov8-instance-segmentation.ipynb)
– not a transformer

[230-yolov8-keypoint-detection.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/230-yolov8-optimization/230-yolov8-keypoint-detection.ipynb)
– not a transformer

[230-yolov8-object-detection.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/230-yolov8-optimization/230-yolov8-object-detection.ipynb)
– not a transformer

[237-segment-anything.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/237-segment-anything/237-segment-anything.ipynb)
– updated ✅

[249-oneformer-segmentation.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/249-oneformer-segmentation/249-oneformer-segmentation.ipynb)
– updated ✅

[255-mms-massively-multilingual-speech.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/255-mms-massively-multilingual-speech/255-mms-massively-multilingual-speech.ipynb)
– updated ✅

[259-decidiffusion-image-generation.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/259-decidiffusion-image-generation/259-decidiffusion-image-generation.ipynb)
– updated ✅

[261-fast-segment-anything.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/261-fast-segment-anything/261-fast-segment-anything.ipynb)
– not a transformer

[263-latent-consistency-models-image-generation.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/263-latent-consistency-models-image-generation/263-latent-consistency-models-image-generation.ipynb)
– updated ✅

[274-efficient-sam.ipynb](https://github.com/openvinotoolkit/openvino_notebooks/blob/main/notebooks/274-efficient-sam/274-efficient-sam.ipynb)
– updated ✅

Ticket 133131

Author: nikita-savelyevv

notebooks/237-segment-anything/237-segment-anything.ipynb

@@ -7,7 +7,8 @@
    "source": [
     "# Object masks from prompts with SAM and OpenVINO\n",
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Background](#Background)\n",
     "- [Prerequisites](#Prerequisites)\n",
     "- [Convert model to OpenVINO Intermediate Representation](#Convert-model-to-OpenVINO-Intermediate-Representation)\n",
@@ -81,7 +82,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "%pip install -q \"segment_anything\" \"gradio>=4.13\" \"openvino>=2023.1.0\" \"nncf>=2.5.0\" \"torch>=2.1\" \"torchvision>=0.16\"  --extra-index-url https://download.pytorch.org/whl/cpu"
+    "%pip install -q \"segment_anything\" \"gradio>=4.13\" \"openvino>=2023.1.0\" \"nncf>=2.7.0\" \"torch>=2.1\" \"torchvision>=0.16\"  --extra-index-url https://download.pytorch.org/whl/cpu"
    ]
   },
   {
@@ -1873,7 +1874,7 @@
     "quantized_model = nncf.quantize(model,\n",
     "                                calibration_dataset,\n",
     "                                model_type=nncf.parameters.ModelType.TRANSFORMER,\n",
-    "                                preset=nncf.common.quantization.structs.QuantizationPreset.MIXED, subset_size=128)\n",
+    "                                subset_size=128)\n",
     "print(\"model quantization finished\")"
    ]
   },
@@ -2238,4 +2239,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

notebooks/249-oneformer-segmentation/249-oneformer-segmentation.ipynb

@@ -28,7 +28,8 @@
    "metadata": {},
    "source": [
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Install required libraries](#Install-required-libraries)\n",
     "- [Prepare the environment](#Prepare-the-environment)\n",
     "- [Load OneFormer fine-tuned on COCO for universal segmentation](#Load-OneFormer-fine-tuned-on-COCO-for-universal-segmentation)\n",
@@ -75,7 +76,7 @@
     }
    ],
    "source": [
-    "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu \"transformers>=4.26.0\" \"openvino>=2023.1.0\" \"nncf>=2.6.0\" gradio torch scipy ipywidgets Pillow matplotlib"
+    "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu \"transformers>=4.26.0\" \"openvino>=2023.1.0\" \"nncf>=2.7.0\" gradio torch scipy ipywidgets Pillow matplotlib"
    ]
   },
   {
@@ -935,7 +936,6 @@
     "        model,\n",
     "        calibration_dataset,\n",
     "        model_type=nncf.parameters.ModelType.TRANSFORMER,\n",
-    "        preset=nncf.QuantizationPreset.MIXED,\n",
     "        subset_size=len(coco_dataset),\n",
     "        # smooth_quant_alpha value of 0.5 was selected based on prediction quality visual examination\n",
     "        advanced_parameters=nncf.AdvancedQuantizationParameters(smooth_quant_alpha=0.5))\n",
@@ -1297,4 +1297,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

notebooks/255-mms-massively-multilingual-speech/255-mms-massively-multilingual-speech.ipynb

@@ -34,7 +34,8 @@
    },
    "source": [
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Prerequisites](#Prerequisites)\n",
     "- [Prepare an example audio](#Prepare-an-example-audio)\n",
     "- [Language Identification (LID)](#Language-Identification-(LID))\n",
@@ -85,7 +86,7 @@
    "outputs": [],
    "source": [
     "%pip install -q --upgrade pip\n",
-    "%pip install -q \"transformers>=4.33.1\" \"openvino>=2023.1.0\" \"numpy>=1.21.0,<=1.24\" \"nncf>=2.6.0\" \n",
+    "%pip install -q \"transformers>=4.33.1\" \"openvino>=2023.1.0\" \"numpy>=1.21.0,<=1.24\" \"nncf>=2.7.0\" \n",
     "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu torch datasets accelerate soundfile librosa gradio jiwer"
    ]
   },
@@ -1300,7 +1301,6 @@
     "    quantized_lid_model = nncf.quantize(\n",
     "        get_lid_model(lid_model_xml_path, compiled=False),\n",
     "        calibration_dataset=nncf.Dataset(calibration_data),\n",
-    "        preset=nncf.QuantizationPreset.MIXED,\n",
     "        subset_size=len(calibration_data),\n",
     "        model_type=nncf.ModelType.TRANSFORMER\n",
     "    )\n",
@@ -1423,7 +1423,6 @@
     "    quantized_asr_model = nncf.quantize(\n",
     "        get_asr_model(asr_model_xml_path_template, language_id, compiled=False),\n",
     "        calibration_dataset=nncf.Dataset(calibration_data),\n",
-    "        preset=nncf.QuantizationPreset.MIXED,\n",
     "        subset_size=len(calibration_data),\n",
     "        model_type=nncf.ModelType.TRANSFORMER\n",
     "    )\n",
@@ -1881,4 +1880,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

notebooks/259-decidiffusion-image-generation/259-decidiffusion-image-generation.ipynb

@@ -23,7 +23,8 @@
     "6. Run Interactive demo for DeciDiffusion model.\n",
     "\n",
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Prerequisites](#Prerequisites)\n",
     "- [Prepare DeciDiffusion models for OpenVINO format conversion](#Prepare-DeciDiffusion-models-for-OpenVINO-format-conversion)\n",
     "    - [About model](#About-model)\n",
@@ -65,7 +66,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu  \"diffusers\" \"transformers\" \"torch\" \"pillow\" \"openvino>=2023.1.0\" \"gradio\" \"datasets\" \"nncf\""
+    "%pip install -q --extra-index-url https://download.pytorch.org/whl/cpu  \"diffusers\" \"transformers\" \"torch\" \"pillow\" \"openvino>=2023.1.0\" \"gradio\" \"datasets\" \"nncf>=2.7.0\""
    ]
   },
   {
@@ -1379,7 +1380,6 @@
     "    quantized_unet = nncf.quantize(\n",
     "        model=unet,\n",
     "        subset_size=subset_size,\n",
-    "        preset=nncf.QuantizationPreset.MIXED,\n",
     "        calibration_dataset=nncf.Dataset(unet_calibration_data),\n",
     "        model_type=nncf.ModelType.TRANSFORMER,\n",
     "        # Smooth Quant algorithm reduces activation quantization error; optimal alpha value was obtained through grid search\n",
@@ -1850,4 +1850,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

notebooks/263-latent-consistency-models-image-generation/263-latent-consistency-models-image-generation.ipynb

@@ -16,7 +16,8 @@
     "In this tutorial, we consider how to convert and run LCM using OpenVINO. An additional part demonstrates how to run quantization with [NNCF](https://github.com/openvinotoolkit/nncf/) to speed up pipeline.\n",
     "\n",
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Prerequisites](#Prerequisites)\n",
     "- [Prepare models for OpenVINO format conversion](#Prepare-models-for-OpenVINO-format-conversion)\n",
     "- [Convert models to OpenVINO format](#Convert-models-to-OpenVINO-format)\n",
@@ -52,7 +53,7 @@
    "outputs": [],
    "source": [
     "%pip install -q \"torch\" --index-url https://download.pytorch.org/whl/cpu\n",
-    "%pip install -q \"openvino>=2023.1.0\" transformers \"diffusers>=0.23.1\" pillow gradio \"nncf>=2.6.0\" datasets --extra-index-url https://download.pytorch.org/whl/cpu"
+    "%pip install -q \"openvino>=2023.1.0\" transformers \"diffusers>=0.23.1\" pillow gradio \"nncf>=2.7.0\" datasets --extra-index-url https://download.pytorch.org/whl/cpu"
    ]
   },
   {
@@ -1261,7 +1262,6 @@
     "    quantized_unet = nncf.quantize(\n",
     "        model=unet,\n",
     "        subset_size=subset_size,\n",
-    "        preset=nncf.QuantizationPreset.MIXED,\n",
     "        calibration_dataset=nncf.Dataset(unet_calibration_data),\n",
     "        model_type=nncf.ModelType.TRANSFORMER,\n",
     "        advanced_parameters=nncf.AdvancedQuantizationParameters(\n",
@@ -2598,4 +2598,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

notebooks/274-efficient-sam/274-efficient-sam.ipynb

@@ -18,7 +18,8 @@
     "\n",
     "\n",
     "\n",
-    "#### Table of contents:\n\n",
+    "#### Table of contents:\n",
+    "\n",
     "- [Prerequisites](#Prerequisites)\n",
     "- [Load PyTorch model](#Load-PyTorch-model)\n",
     "- [Run PyTorch model inference](#Run-PyTorch-model-inference)\n",
@@ -57,7 +58,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "%pip install -q \"openvino>=2023.3.0\" \"nncf>=2.7\" opencv-python matplotlib \"gradio>=4.13\" torch torchvision --extra-index-url https://download.pytorch.org/whl/cpu"
+    "%pip install -q \"openvino>=2023.3.0\" \"nncf>=2.7.0\" opencv-python matplotlib \"gradio>=4.13\" torch torchvision --extra-index-url https://download.pytorch.org/whl/cpu"
    ]
   },
   {
@@ -933,7 +934,7 @@
     "quantized_model = nncf.quantize(model,\n",
     "                                calibration_dataset,\n",
     "                                model_type=nncf.parameters.ModelType.TRANSFORMER,\n",
-    "                                preset=nncf.common.quantization.structs.QuantizationPreset.MIXED, subset_size=128)\n",
+    "                                subset_size=128)\n",
     "print(\"model quantization finished\")"
    ]
   },
@@ -1837,4 +1838,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}