diff --git a/FAQ.md b/FAQ.md index 6f6c9c441..c44e40e4f 100644 --- a/FAQ.md +++ b/FAQ.md @@ -36,7 +36,7 @@ quantized_model = mct.keras_load_quantized_model('my_model.keras') #### PyTorch -PyTorch models can be exported as onnx models. An example of loading a saved onnx model can be found [here](https://sony.github.io/model_optimization/api/api_docs/modules/exporter.html#use-exported-model-for-inference). +PyTorch models can be exported as onnx models. An example of loading a saved onnx model can be found [here](https://github.com/sony/model_optimization/blob/main/docs/api/experimental_api_docs/modules/exporter.html#use-exported-model-for-inference). *Note:* Running inference on an ONNX model in the `onnxruntime` package has a high latency. Inference on the target platform (e.g. the IMX500) is not affected by this latency. diff --git a/README.md b/README.md index f20bc3531..02e59a144 100644 --- a/README.md +++ b/README.md @@ -38,15 +38,16 @@ For installing the nightly version or installing from source, refer to the [inst ### Quick start & tutorials -For an example of how to use MCT with TensorFlow or PyTorch on various models and tasks, -check out the [quick-start page](tutorials/quick_start/README.md) and -the [results CSV](tutorials/quick_start/results/model_quantization_results.csv). - -In addition, a set of [notebooks](tutorials/notebooks) are provided for an easy start. For example: -* [MobileNet with Tensorflow](tutorials/notebooks/keras/ptq/example_keras_mobilenet.py). -* [MobileNetV2 with PyTorch](tutorials/notebooks/pytorch/ptq/example_pytorch_mobilenet_v2.py). +Explore the Model Compression Toolkit (MCT) through our tutorials, +covering compression techniques for Keras and PyTorch models. Access interactive [notebooks](tutorials/README.md) +for hands-on learning. For example: +* [Keras MobileNetV2 post training quantization](tutorials/notebooks/keras/ptq/example_keras_imagenet.ipynb) +* [Post training quantization with PyTorch](tutorials/notebooks/pytorch/ptq/example_pytorch_quantization_mnist.ipynb) * [Data Generation for ResNet18 with PyTorch](tutorials/notebooks/pytorch/data_generation/example_pytorch_data_generation.ipynb). +Additionally, for quick quantization of a variety of models from well-known collections, +visit the [quick-start page](tutorials/quick_start/README.md) and the +[results CSV](tutorials/quick_start/results/model_quantization_results.csv). ### Supported Versions diff --git a/tutorials/README.md b/tutorials/README.md index 42decc234..ddfe990c3 100644 --- a/tutorials/README.md +++ b/tutorials/README.md @@ -6,8 +6,8 @@ Access interactive Jupyter notebooks for hands-on learning. ## Getting started Learn how to quickly quantize pre-trained models using MCT's post-training quantization technique for both Keras and PyTorch models. -- [Keras MobileNetV2 post training quantization](notebooks/keras/ptq/example_keras_imagenet.ipynb) -- [Pytorch MobileNetV2 post training quantization](notebooks/pytorch/ptq/example_pytorch_quantization_mnist.ipynb) +- [Post training quantization with Keras](notebooks/keras/ptq/example_keras_imagenet.ipynb) +- [Post training quantization with PyTorch](notebooks/pytorch/ptq/example_pytorch_quantization_mnist.ipynb) ## MCT Features This set of tutorials covers all the quantization tools provided by MCT. @@ -15,7 +15,7 @@ The notebooks in this section demonstrate how to configure and run simple and ad This includes fine-tuning PTQ (Post-Training Quantization) configurations, exporting models, and exploring advanced compression techniques. These techniques are essential for further optimizing models and achieving superior performance in deployment scenarios. -- [MCT notebooks](notebooks/MCT_notebooks.md) +- [MCT Features notebooks](notebooks/MCT_notebooks.md) ## Quantization for Sony-IMX500 deployment @@ -23,4 +23,4 @@ This section provides several guides on quantizing pre-trained models to meet sp [Sony-IMX500](https://developer.sony.com/imx500/) processing platform. We will cover various tasks and demonstrate the necessary steps to achieve efficient quantization for optimal deployment performance. -- [IMX500 notebooks](notebooks/IMX500_notebooks.md) +- [MCT IMX500 notebooks](notebooks/IMX500_notebooks.md) diff --git a/tutorials/notebooks/IMX500_notebooks.md b/tutorials/notebooks/IMX500_notebooks.md index 2495497d3..ad914988c 100644 --- a/tutorials/notebooks/IMX500_notebooks.md +++ b/tutorials/notebooks/IMX500_notebooks.md @@ -6,7 +6,7 @@ deployment performance. | Task | Model | Source Repository | Notebook | |-----------------------------------------------------------------|----------------|---------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------| - | Classification | MobileNetV2 | [Keras Applications](https://keras.io/api/applications/) | [Keras notebook](model_optimization/tutorials/notebooks/keras/ptq/example_keras_imagenet.ipynb) | - | Object Detection | YOLOv8n | [Ultralytics](https://github.com/ultralytics/ultralytics) | [Keras notebook](model_optimization/tutorials/notebooks/keras/ptq/keras_yolov8n_for_imx500.ipynb) | - | Semantic Segmentation | DeepLabV3-Plus | [bonlime's repo](https://github.com/bonlime/keras-deeplab-v3-plus) | [Keras notebook](model_optimization/tutorials/notebooks/keras/ptq/keras_deeplabv3plus_for_imx500.ipynb) | + | Classification | MobileNetV2 | [Keras Applications](https://keras.io/api/applications/) | [Keras notebook](keras/ptq/example_keras_imagenet.ipynb) | + | Object Detection | YOLOv8n | [Ultralytics](https://github.com/ultralytics/ultralytics) | [Keras notebook](keras/ptq/keras_yolov8n_for_imx500.ipynb) | + | Semantic Segmentation | DeepLabV3-Plus | [bonlime's repo](https://github.com/bonlime/keras-deeplab-v3-plus) | [Keras notebook](keras/ptq/keras_deeplabv3plus_for_imx500.ipynb) | diff --git a/tutorials/notebooks/MCT_notebooks.md b/tutorials/notebooks/MCT_notebooks.md index 44a4050ae..251194ad8 100644 --- a/tutorials/notebooks/MCT_notebooks.md +++ b/tutorials/notebooks/MCT_notebooks.md @@ -92,14 +92,6 @@ These techniques are essential for further optimizing models and achieving super -
- Quantization-Aware Training (QAT) - - | Tutorial | Included Features | - |-----------------------------------------------------------------------------------|--------------| - | [QAT on MNIST](pytorch/qat/example_pytorch_qat.py) | ✅ QAT | -
-