English | 简体中文

📚 OpenVINO™ Notebooks

A collection of ready-to-run Jupyter notebooks for learning and experimenting with the OpenVINO™ Toolkit. The notebooks provide an introduction to OpenVINO basics and teach developers how to leverage our API for optimized deep learning inference.

🚀 AI Trends - Notebooks

Check out the latest notebooks that show how to optimize and deploy popular models on Intel CPU and GPU.

Notebook	Description	Preview	Complementary Materials
YOLOv8 - Optimization: Object Detection Instance Segmentation Keypoints Detection	Optimize YOLOv8 using NNCF PTQ API		Blog - How to get YOLOv8 Over 1000 fps with Intel GPUs?
SAM - Segment Anything Model	Prompt based object segmentation mask generation using Segment Anything and OpenVINO™		Blog - SAM: Segment Anything Model — Versatile by itself and Faster by OpenVINO
ControlNet - Stable-Diffusion	A Text-to-Image Generation with ControlNet Conditioning and OpenVINO™		Blog - Control your Stable Diffusion Model with ControlNet and OpenVINO
Stable Diffusion v2	Text-to-Image Generation and Infinite Zoom with Stable Diffusion v2 and OpenVINO™		Blog - How to run Stable Diffusion on Intel GPUs with OpenVINO
Whisper - Subtitles generation	Generate subtitles for video with OpenAI Whisper and OpenVINO
CLIP - zero-shot-image-classification	Perform Zero-shot Image Classification with CLIP and OpenVINO		Blog - Generative AI and Explainable AI with OpenVINO
BLIP - Visual-language-processing	Visual Question Answering and Image Captioning using BLIP and OpenVINO™		Blog - Multimodality with OpenVINO — BLIP
Instruct pix2pix - Image-editing	Image editing with InstructPix2Pix		Blog - Generative AI and Explainable AI with OpenVINO
DeepFloyd IF - Text-to-Image generation	Text-to-Image Generation with DeepFloyd IF and OpenVINO™
ImageBind	Binding multimodal data using ImageBind and OpenVINO™
Dolly v2	Instruction following using Databricks Dolly 2.0 and OpenVINO™
Stable Diffusion XL and Segmind Stable Diffusion 1B (SSD-1B)	Image generation with Stable Diffusion XL and Segmind Stable Diffusion 1B (SSD-1B) and OpenVINO™
MusicGen	Controllable Music Generation with MusicGen and OpenVINO™
Tiny SD	Image Generation with Tiny-SD and OpenVINO™
ZeroScope Text-to-video synthesis	Text-to-video synthesis with ZeroScope and OpenVINO™	A panda eating bamboo on a rock
LLM chatbot	Create LLM-powered Chatbot using OpenVINO™
QA over Document	Create LLM-powered RAG system using OpenVINO™ and LangChain
Bark Text-to-Speech	Text-to-Speech generation using Bark and OpenVINO™
LLaVA Multimodal Chatbot	Visual-language assistant with LLaVA and OpenVINO™
BLIP-Diffusion - Subject-Driven Generation	Subject-driven image generation and editing using BLIP Diffusion and OpenVINO™
DeciDiffusion	Image generation with DeciDiffusion and OpenVINO™
Fast Segment Anything	Object segmentations with FastSAM and OpenVINO™
SoftVC VITS Singing Voice Conversion	SoftVC VITS Singing Voice Conversion and OpenVINO™
Latent Consistency Models: the next generation of Image Generation models	Image generation with Latent Consistency Models (LCM) and OpenVINO™
Speedup ControlNet pipeline with LCM LoRA	Text-to-Image Generation with LCM LoRA and ControlNet Conditioning
QR Code Monster	Generate creative QR codes with ControlNet QR Code Monster and OpenVINO™
Würstchen	Text-to-image generation with Würstchen and OpenVINO™
Distil-Whisper	Automatic speech recognition using Distil-Whisper and OpenVINO™
FILM	Frame interpolation with FILM and OpenVINO™
Audio LDM 2	Sound Generation with AudioLDM2 and OpenVINO™
SDXL-Turbo	Single-step image generation using SDXL-turbo and OpenVINO
Stable-Zephyr chatbot	Use Stable-Zephyr as chatbot assistant with OpenVINO
Efficient-SAM	Object segmentation with EfficientSAM and OpenVINO
LLM Instruction following pipeline	Usage variety of LLM models for answering questions using OpenVINO
Stable Diffusion with IP-Adapter	Image conditioning in Stable Diffusion pipeline using IP-Adapter
MobileVLM	Mobile language assistant with MobileVLM and OpenVINO
DepthAnything	Monocular Depth estimation with DepthAnything and OpenVINO
Kosmos-2: Grounding Multimodal Large Language Models	Kosmos-2: Grounding Multimodal Large Language Model and OpenVINO™
PhotoMaker	Text-to-image generation using PhotoMaker and OpenVINO
OpenVoice	OpenVoice a versatile instant voice tone transferring and generating speech in various languages.
InstantID	InstantID: Zero-shot Identity-Preserving Image Generation using OpenVINO

Table of Contents

• 🚀 AI Trends - Notebooks
• Table of Contents
• 📝 Installation Guide
• 🚀 Getting Started
  • 💻 First steps
  • ⌚ Convert & Optimize
  • 🎯 Model Demos
  • 🏃 Model Training
  • 📺 Live Demos
• ⚙️ System Requirements
• 💻 Run the Notebooks
  • To Launch a Single Notebook
  • To Launch all Notebooks
• 🧹 Cleaning Up
• ⚠️ Troubleshooting
• 🧑‍💻 Contributors
• ❓ FAQ

📝 Installation Guide

OpenVINO Notebooks require Python and Git. To get started, select the guide for your operating system or environment:

Windows	Ubuntu	macOS	Red Hat	CentOS	Azure ML	Docker	Amazon SageMaker

🚀 Getting Started

The Jupyter notebooks are categorized into four classes, select one related to your needs or give them all a try. Good Luck!

NOTE: The main branch of this repository was updated to support the new OpenVINO 2023.3 release. To upgrade to the new release version, please run pip install --upgrade -r requirements.txt in your openvino_env virtual environment. If you need to install for the first time, see the Installation Guide section below. If you wish to use the previous release version of OpenVINO, please checkout the 2023.2 branch. If you wish to use the previous Long Term Support (LTS) version of OpenVINO check out the 2022.3 branch.

If you need help, please start a GitHub Discussion.

💻 First steps

Brief tutorials that demonstrate how to use OpenVINO's Python API for inference.

001-hello-world	002-openvino-api	003-hello-segmentation	004-hello-detection
Classify an image with OpenVINO	Learn the OpenVINO Python API	Semantic segmentation with OpenVINO	Text detection with OpenVINO

⌚ Convert & Optimize

Tutorials that explain how to optimize and quantize models with OpenVINO tools.

Notebook	Description
101-tensorflow-classification-to-openvino	Convert TensorFlow models to OpenVINO IR
102-pytorch-to-openvino	Convert PyTorch models to OpenVINO IR
103-paddle-to-openvino	Convert PaddlePaddle models to OpenVINO IR
104-model-tools	Download, convert and benchmark models from Open Model Zoo
105-language-quantize-bert	Optimize and quantize a pre-trained BERT model
106-auto-device	Demonstrate how to use AUTO Device
107-speech-recognition-quantization	Quantize speech recognition models using NNCF PTQ API
108-gpu-device	Working with GPUs in OpenVINO™
109-performance-tricks	Performance tricks in OpenVINO™
110-ct-segmentation-quantize	Quantize a kidney segmentation model and show live inference
112-pytorch-post-training-quantization-nncf	Use Neural Network Compression Framework (NNCF) to quantize PyTorch model in post-training mode (without model fine-tuning)
113-image-classification-quantization	Quantize Image Classification model
115-async-api	Use Asynchronous Execution to Improve Data Pipelining
116-sparsity-optimization	Improve performance of sparse Transformer models
117-model-server	Introduction to model serving with OpenVINO™ Model Server (OVMS)
118-optimize-preprocessing	Improve performance of image preprocessing step
119-tflite-to-openvino	Convert TensorFlow Lite models to OpenVINO IR
120-tensorflow-object-detection-to-openvino	Convert TensorFlow Object Detection models to OpenVINO IR
121-convert-to-openvino	Learn OpenVINO model conversion API
122-quantizing-model-with-accuracy-control	Quantizing with Accuracy Control using NNCF
123-detectron2-to-openvino	Convert Detectron2 models to OpenVINO IR
124-hugging-face-hub	Load models from Hugging Face Model Hub with OpenVINO™
125-torchvision-zoo-to-openvino Classification Semantic Segmentation	Convert torchvision classification and semantic segmentation models to OpenVINO IR
126-tensorflow-hub	Convert TensorFlow Hub models to OpenVINO IR
127-big-transfer-quantization	BiT Image Classification OpenVINO IR model Quantization with NNCF

🎯 Model Demos

Demos that demonstrate inference on a particular model.

Notebook	Description	Preview
201-vision-monodepth	Monocular depth estimation with images and video
202-vision-superresolution-image	Upscale raw images with a super resolution model	→
202-vision-superresolution-video	Turn 360p into 1080p video using a super resolution model	→
203-meter-reader	PaddlePaddle pre-trained models to read industrial meter's value
204-segmenter-semantic-segmentation	Semantic Segmentation with OpenVINO™ using Segmenter
205-vision-background-removal	Remove and replace the background in an image using salient object detection
206-vision-paddlegan-anime	Turn an image into anime using a GAN	→
207-vision-paddlegan-superresolution	Upscale small images with superresolution using a PaddleGAN model
208-optical-character-recognition	Annotate text on images using text recognition resnet
209-handwritten-ocr	OCR for handwritten simplified Chinese and Japanese	的人不一了是他有为在责新中任自之我们
210-slowfast-video-recognition	Video Recognition using SlowFast and OpenVINO™
211-speech-to-text	Run inference on speech-to-text recognition model
212-pyannote-speaker-diarization	Run inference on speaker diarization pipeline
213-question-answering	Answer your questions basing on a context
214-grammar-correction	Grammatical Error Correction with OpenVINO	Input text: I'm working in campany for last 2 yeas. Generated text: I'm working in a company for the last 2 years.
215-image-inpainting	Fill missing pixels with image in-painting
216-attention-center	The attention center model with OpenVINO™
218-vehicle-detection-and-recognition	Use pre-trained models to detect and recognize vehicles and their attributes with OpenVINO
219-knowledge-graphs-conve	Optimize the knowledge graph embeddings model (ConvE) with OpenVINO
220-books-alignment-labse	Cross-lingual Books Alignment With Transformers and OpenVINO™
221-machine-translation	Real-time translation from English to German
222-vision-image-colorization	Use pre-trained models to colorize black & white images using OpenVINO
223-text-prediction	Use pretrained models to perform text prediction on an input sequence
224-3D-segmentation-point-clouds	Process point cloud data and run 3D Part Segmentation with OpenVINO
225-stable-diffusion-text-to-image	Text-to-image generation with Stable Diffusion method
226-yolov7-optimization	Optimize YOLOv7 using NNCF PTQ API
227-whisper-subtitles-generation	Generate subtitles for video with OpenAI Whisper and OpenVINO
228-clip-zero-shot-image-classification	Perform Zero-shot Image Classification with CLIP and OpenVINO
229-distilbert-sequence-classification	Sequence Classification with OpenVINO
230-yolov8-object-detection	Optimize YOLOv8 object detection using NNCF PTQ API
230-yolov8-instance-segmentation	Optimize YOLOv8 instance segmentation using NNCF PTQ API
230-yolov8-keypoint-detection	Optimize YOLOv8 keypoint detection using NNCF PTQ API
231-instruct-pix2pix-image-editing	Image editing with InstructPix2Pix
232-clip-language-saliency-map	Language-Visual Saliency with CLIP and OpenVINO™
233-blip-visual-language-processing	Visual Question Answering and Image Captioning using BLIP and OpenVINO™
234-encodec-audio-compression	Audio compression with EnCodec and OpenVINO™
235-controlnet-stable-diffusion	A Text-to-Image Generation with ControlNet Conditioning and OpenVINO™
236-stable-diffusion-v2	Text-to-Image Generation and Infinite Zoom with Stable Diffusion v2 and OpenVINO™
237-segment-anything	Prompt based segmentation using Segment Anything and OpenVINO™.
238-deep-floyd-if	Text-to-Image Generation with DeepFloyd IF and OpenVINO™
239-image-bind	Binding multimodal data using ImageBind and OpenVINO™
240-dolly-2-instruction-following	Instruction following using Databricks Dolly 2.0 and OpenVINO™
241-riffusion-text-to-music	Text-to-Music generation using Riffusion and OpenVINO™
242-freevc-voice-conversion	High-Quality Text-Free One-Shot Voice Conversion with FreeVC and OpenVINO™
243-tflite-selfie-segmentation	Selfie Segmentation using TFLite and OpenVINO™
244-named-entity-recognition	Named entity recognition with OpenVINO™
245-typo-detector	English Typo Detection in sentences with OpenVINO™
246-depth-estimation-videpth	Monocular Visual-Inertial Depth Estimation with OpenVINO™
247-code-language-id	Identify the programming language used in an arbitrary code snippet
248-stable-diffusion-xl	Image generation with Stable Diffusion XL and OpenVINO™
249-oneformer-segmentation	Universal segmentation with OneFormer and OpenVINO™
250-music-generation	Controllable Music Generation with MusicGen and OpenVINO™
251-tiny-sd-image-generation	Image Generation with Tiny-SD and OpenVINO™
252-fastcomposer-image-generation	Image generation with FastComposer and OpenVINO™
253-zeroscope-text2video	Text-to-video synthesis with ZeroScope and OpenVINO™	A panda eating bamboo on a rock
254-llm-chatbot	Create LLM-powered Chatbot using OpenVINO™
255-mms-massively-multilingual-speech	MMS: Scaling Speech Technology to 1000+ languages with OpenVINO™
256-bark-text-to-audio	Text-to-Speech generation using Bark and OpenVINO™
257-llava-multimodal-chatbot	Visual-language assistant with LLaVA and OpenVINO™
258-blip-diffusion-subject-generation	Subject-driven image generation and editing using BLIP Diffusion and OpenVINO™
259-decidiffusion-image-generation	Image generation with DeciDiffusion and OpenVINO™
260-pix2struct-docvqa	Document Visual Question Answering using Pix2Struct and OpenVINO™
261-fast-segment-anything	Object segmentations with FastSAM and OpenVINO™
262-softvc-voice-conversion	SoftVC VITS Singing Voice Conversion and OpenVINO™
263-latent-consistency-models-image-generation	Image generation with Latent Consistency Models (LCM) and OpenVINO™
264-qrcode-monster	Generate creative QR codes with ControlNet QR Code Monster and OpenVINO™
265-wuerstchen-image-generation	Text-to-image generation with Würstchen and OpenVINO™
266-speculative-sampling	Text Generation via Speculative Sampling, KV Caching, and OpenVINO™
267-distil-whisper-asr	Automatic speech recognition using Distil-Whisper and OpenVINO™
268-table-question-answering	Table Question Answering using TAPAS and OpenVINO™
269-film-slowmo	Frame interpolation with FILM and OpenVINO™
270-sound-generation-audioldm2	Sound Generation with AudioLDM2 and OpenVINO™
271-sdxl-turbo	Single-step image generation using SDXL-turbo and OpenVINO
272-paint-by-example	Exemplar based image editing using diffusion models, Paint-by-Example, and OpenVINO™
273-stable-zephyr-3b-chatbot	Use Stable-Zephyr as chatbot assistant with OpenVINO
274-efficient-sam	Object segmentation with EfficientSAM and OpenVINO™
275-llm-question-answering	LLM Instruction following pipeline
276-stable-diffusion-torchdynamo-backend	Image generation with Stable Diffusion and OpenVINO™ torch.compile feature
277-amused-lightweight-text-to-image	Lightweight image generation with aMUSEd and OpenVINO™
278-stable-diffusion-ip-adapter	Image conditioning in Stable Diffusion pipeline using IP-Adapter
279-mobilevlm-language-assistant	Mobile language assistant with MobileVLM and OpenVINO
280-depth-anything	Monocular Depth Estimation with DepthAnything and OpenVINO
281-kosmos2-multimodal-large-language-model	Kosmos-2: Multimodal Large Language Model and OpenVINO™
282-siglip-zero-shot-image-classification	Zero-shot Image Classification with SigLIP
283-photo-maker	Text-to-image generation using PhotoMaker and OpenVINO
284-openvoice	OpenVoice a versatile instant voice tone transferring and generating speech in various languages.
285-surya-line-level-text-detection	Line-level text detection with Surya
286-instant-id	InstantID: Zero-shot Identity-Preserving Image Generation using OpenVINO

🏃 Model Training

Tutorials that include code to train neural networks.

Notebook	Description	Preview
301-tensorflow-training-openvino	Train a flower classification model from TensorFlow, then convert to OpenVINO IR
301-tensorflow-training-openvino-nncf	Use Neural Network Compression Framework (NNCF) to quantize model from TensorFlow
302-pytorch-quantization-aware-training	Use Neural Network Compression Framework (NNCF) to quantize PyTorch model
305-tensorflow-quantization-aware-training	Use Neural Network Compression Framework (NNCF) to quantize TensorFlow model

📺 Live Demos

Live inference demos that run on a webcam or video files.

Notebook	Description	Preview
401-object-detection-webcam	Object detection with a webcam or video file
402-pose-estimation-webcam	Human pose estimation with a webcam or video file
403-action-recognition-webcam	Human action recognition with a webcam or video file
404-style-transfer-webcam	Style Transfer with a webcam or video file
405-paddle-ocr-webcam	OCR with a webcam or video file
406-3D-pose-estimation-webcam	3D display of human pose estimation with a webcam or video file
407-person-tracking-webcam	Person tracking with a webcam or video file

If you run into issues, please check the troubleshooting section, FAQs or start a GitHub discussion.

Notebooks with and buttons can be run without installing anything. Binder and Google Colab are free online services with limited resources. For the best performance, please follow the Installation Guide and run the notebooks locally.

⚙️ System Requirements

The notebooks run almost anywhere — your laptop, a cloud VM, or even a Docker container. The table below lists the supported operating systems and Python versions.

Supported Operating System	Python Version (64-bit)
Ubuntu 20.04 LTS, 64-bit	3.8 - 3.10
Ubuntu 22.04 LTS, 64-bit	3.8 - 3.10
Red Hat Enterprise Linux 8, 64-bit	3.8 - 3.10
CentOS 7, 64-bit	3.8 - 3.10
macOS 10.15.x versions or higher	3.8 - 3.10
Windows 10, 64-bit Pro, Enterprise or Education editions	3.8 - 3.10
Windows Server 2016 or higher	3.8 - 3.10

💻 Run the Notebooks

To Launch a Single Notebook

If you wish to launch only one notebook, like the Monodepth notebook, run the command below.

jupyter 201-vision-monodepth.ipynb

To Launch all Notebooks

jupyter lab notebooks

In your browser, select a notebook from the file browser in Jupyter Lab using the left sidebar. Each tutorial is located in a subdirectory within the notebooks directory.

🧹 Cleaning Up

1. Shut Down Jupyter Kernel

   To end your Jupyter session, press Ctrl-c. This will prompt you to Shutdown this Jupyter server (y/[n])? enter y and hit Enter.

2. Deactivate Virtual Environment

   To deactivate your virtualenv, simply run deactivate from the terminal window where you activated openvino_env. This will deactivate your environment.

   To reactivate your environment, run source openvino_env/bin/activate on Linux or openvino_env\Scripts\activate on Windows, then type jupyter lab or jupyter notebook to launch the notebooks again.

3. Delete Virtual Environment (Optional)

   To remove your virtual environment, simply delete the openvino_env directory:

• On Linux and macOS:

  rm -rf openvino_env

• On Windows:

  rmdir /s openvino_env

• Remove openvino_env Kernel from Jupyter

  jupyter kernelspec remove openvino_env

⚠️ Troubleshooting

If these tips do not solve your problem, please open a discussion topic or create an issue!

• To check some common installation problems, run python check_install.py. This script is located in the openvino_notebooks directory. Please run it after activating the openvino_env virtual environment.
• If you get an ImportError, double-check that you installed the Jupyter kernel. If necessary, choose the openvino_env kernel from the Kernel->Change Kernel menu in Jupyter Lab or Jupyter Notebook.
• If OpenVINO is installed globally, do not run installation commands in a terminal where setupvars.bat or setupvars.sh are sourced.
• For Windows installation, it is recommended to use Command Prompt (cmd.exe), not PowerShell.

🧑‍💻 Contributors

Made with contrib.rocks.

❓ FAQ

• Which devices does OpenVINO support?
• What is the first CPU generation you support with OpenVINO?
• Are there any success stories about deploying real-world solutions with OpenVINO?

---

* Other names and brands may be claimed as the property of others.