The Intel AI Tools give data scientists, AI developers, and researchers familiar Python* tools and frameworks to accelerate end-to-end data science and analytics pipelines on Intel® architectures. The components are built using oneAPI libraries for low-level compute optimizations. This toolkit maximizes performance from preprocessing through machine learning, and provides interoperability for efficient model development.
You can find more information at AI Tools.
Users can explore the extensive features of Intel AI Tools through provided feature and functionality samples, offering a deeper understanding of their capabilities.
Code samples are licensed under the MIT license. See License.txt for details.
Third party program Licenses can be found here: third-party-programs.txt
| AI Tools preset bundle | Component | Folder | Description |
|---|---|---|---|
| Deep Learning | Intel® Neural Compressor (INC) | Intel® Neural Compressor (INC) Quantization Aware Training | Fine-tune a BERT tiny model for emotion classification task using Quantization Aware Training and Inference from Intel® Neural Compressor (INC). |
| Deep Learning | Intel® Extension for PyTorch (IPEX) | IntelPyTorch Extensions Inference Optimization | Apply Intel® Extension for PyTorch (IPEX) to a PyTorch workload to gain performance boost. |
| Deep Learning | Intel® Extension for PyTorch (IPEX) | IntelPyTorch Extensions GPU Inference Optimization with AMP | Use the PyTorch ResNet50 model transfer learning and inference using the CIFAR10 dataset on Intel discrete GPU with Intel® Extension for PyTorch (IPEX). |
| Deep Learning | Intel® Extension for PyTorch (IPEX) | IntelPyTorch_InferenceOptimizations_AMX_BF16_INT8 | Analyze inference performance improvements using Intel® Extension for PyTorch (IPEX) with Advanced Matrix Extensions (Intel® AMX) Bfloat16 and Integer8. |
| Deep Learning | PyTorch | IntelPyTorch TrainingOptimizations Intel® AMX BF16 | Analyze training performance improvements using Intel® Extension for PyTorch (IPEX) with Intel® AMX Bfloat16. |
| Data Analytics | Numpy, Numba | IntelPython Numpy Numba dpex kNN | Optimize k-NN model by numba_dpex operations without sacrificing accuracy. |
| Classical Machine Learning | XGBoost | IntelPython XGBoost Performance | Analyze the performance benefit from using Intel optimized XGBoost compared to un-optimized XGBoost 0.81. |
| Classical Machine Learning | XGBoost | IntelPython XGBoost daal4pyPrediction | Analyze the performance benefit of minimal code changes to port pre-trained XGBoost model to daal4py prediction for much faster prediction than XGBoost prediction. |
| Classical Machine Learning | daal4py | IntelPython daal4py DistributedKMeans | Train and predict with a distributed k-means model using the python API package daal4py powered by the oneAPI Data Analytics Library. |
| Classical Machine Learning | daal4py | IntelPython daal4py DistributedLinearRegression | Run a distributed Linear Regression model with oneAPI Data Analytics Library (oneDAL) daal4py library memory objects. |
| Deep Learning | PyTorch | IntelPytorch Interactive Chat Quantization | Create interactive chat based on pre-trained DialoGPT model and add the Intel® Extension for PyTorch (IPEX) quantization to it. |
| Deep Learning | PyTorch | IntelPytorch Quantization | Inference performance improvements using Intel® Extension for PyTorch (IPEX) with feature quantization. |
| Deep Learning | TensorFlow | IntelTensorFlow Intel® AMX BF16 Training | Enabling auto-mixed precision to use low-precision datatypes, like bfloat16, for model inference with TensorFlow* . |
| Deep Learning | TensorFlow | IntelTensorFlow Intel® AMX BF16 Training | Training performance improvements with Intel® AMX BF16. |
| Deep Learning | TensorFlow | IntelTensorFlow Enabling Auto Mixed Precision for TransferLearning | Enabling auto-mixed precision to use low-precision datatypes, like bfloat16, for transfer learning with TensorFlow*. |
| Deep Learning | Horovod | IntelTensorFlow Horovod Distributed Deep Learning | Run inference & training workloads across multi-cards using Intel Optimization for Horovod and TensorFlow* on Intel® dGPU's. |
| Deep Learning | TensorFlow | IntelTensorFlow InferenceOptimization | Optimize a pre-trained model for a better inference performance. |
| Deep Learning | TensorFlow & Intel® AI Reference Models | IntelTensorFlow Reference Models Inference with FP32 Int8 | Run ResNet50 inference on Intel's pretrained FP32 and NT8 model. |
| Deep Learning | TensorFlow | IntelTensorFlow PerformanceAnalysis | Analyze the performance difference between Stock Tensorflow and Intel Tensorflow. |
| Deep Learning | TensorFlow | IntelTensorFlow Transformer Intel® AMX bfloat16 MixedPrecisiong | Run a transformer classification model with bfloat16 mixed precision. |
| Deep Learning | TensorFlow | IntelTensorFlow for LLMs | Finetune a GPT-J (LLM) model using the GLUE cola dataset with the Intel® Optimization for TensorFlow*. |
| Classical Machine Learning | Scikit-learn | IntelScikitLearn Extensions SVC Adult | Use Intel® Extension for Scikit-learn to accelerate the training and prediction with SVC algorithm on Adult dataset. Compare the performance of SVC algorithm optimized through Intel® Extension for Scikit-learn against original Scikit-learn. |
*Other names and brands may be claimed as the property of others. Trademarks