This page provides overview of AI Containers for many Intel optimized open-source frameworks such as PyTorch, TensorFlow, SciKit Learn, XGBoost, Modin etc. Also provides containers for open-source deep learning models optimized by Intel to run on Intel® Xeon® Scalable processors and Intel® Data Center GPUs.
Docker container with Intel® Distribution of Python
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Intel® Distribution of Python - Full | - | docker pull intel/python:py310-full |
1.38GB |
| Intel® Distribution of Python - Core | - | docker pull intel/python:py310-core |
814.77MB |
Base containers to run on Intel® Xeon® Scalable processors
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Intel® Optimizations for TensorFlow (PIP) | TensorFlow | docker pull intel/intel-optimized-tensorflow:2.14.0-pip-base |
405.53MB |
| Intel® Optimizations for TensorFlow with Jupyter notebook support (PIP) | TensorFlow | docker pull intel/intel-optimized-tensorflow:2.14.0-pip-jupyter |
464MB |
| Intel® Optimizations for TensorFlow Multi-node with OpenMPI (PIP) | TensorFlow | docker pull intel/intel-optimized-tensorflow:2.14.0-pip-openmpi-multinode |
817MB |
Intel® Data Center GPU Max Series and Intel® Data Center GPU Flex Series base containers
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Intel® Extension for PyTorch* Container for Flex Series | PyTorch | docker pull intel-extension-for-pytorch:xpu-flex |
7.04GB |
| Intel® Extension for PyTorch* Container for Max Series | PyTorch | docker pull intel-extension-for-pytorch:xpu-max |
7.13GB |
| Intel® Extension for TensorFlow | TensoFlow | docker pull intel/intel-extension-for-tensorflow:xpu |
2.29GB |
These are the deep learning Models optimized to run on Intel® Xeon® Scalable processors
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| ResNet 50v1.5 Training | TensorFlow | docker pull intel/image-recognition:centos-tf-cpu-resnet50v1-5-training |
1.46GB |
| ResNet 50v1.5 Inference | TensorFlow | docker pull intel/image-recognition:centos-tf-cpu-resnet50v1-5-inference |
1.69GB |
| BERT large Training | TensorFlow | docker pull intel/language-modeling:centos-tf-cpu-bert-large-pretraining |
1.46GB |
These are the deep learning Models optimized to run on Intel's Flex Series GPU platform.
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| ResNet50 v1.5 Inference | TensorFlow | docker pull intel/image-recognition:tf-flex-gpu-resnet50v1-5-inference |
2.35GB |
| MaskRCNN Inference | TensorFlow | docker pull intel/image-segmentation:tf-flex-gpu-maskrcnn-inference |
2.59GB |
| EfficientNet Inference | TensorFlow | docker pull intel/image-recognition:tf-flex-gpu-efficientnet-inference |
2.3GB |
These are the deep learning Models optimized to run on Intel's Max Series GPU platform.
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| ResNet50 v1.5 Inference | PyTorch | docker pull intel/image-recognition:pytorch-max-gpu-resnet50v1-5-inference |
2.37GB |
| ResNet50 v1.5 Training | PyTorch | docker pull intel/image-recognition:pytorch-max-gpu-resnet50v1-5-training |
2.51GB |
| BERT Large Inference | PyTorch | docker pull intel/language-modeling:pytorch-max-gpu-bert-large-inference |
7.8GB |
These are Domain Toolkit Containers that use the latest transformers and LLM technology with Intel optimizations.
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Intel® Transfer Learning Toolkit | PyTorch & TensorFlow | docker pull intel/ai-tools:tlt-0.5.0 |
910.64MB |
| Intel® Extension for Transformers Chatbot | PyTorch & TensorFlow | docker pull intel/ai-tools:itrex-chatbot |
4.54GB |
| Intel® Extension for Transformers | PyTorch & TensorFlow | docker pull intel/ai-tools:itrex-1.3.0 |
1.04GB |
Intel® AI Tools Selector Preset Containers provides development environment with Intel optimized software for data scientists and developers.
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Data Analytics py3.9 | - | docker pull intel/data-analytics:latest-py3.9 |
1.97GB |
| Data Analytics py3.10 | - | docker pull intel/data-analytics:latest-py3.10 |
1.98GB |
| Classical ML py3.9 | - | docker pull intel/classical-ml:latest-py3.9 |
2.11GB |
These containers are for domain use-cases like Disease Prediction, Credit Card Fraud Detection, and Document Automation.
| Container | Framework | Docker Pull Command | Compressed Size |
|---|---|---|---|
| Document Automation Fine Tuning | - | docker pull intel/ai-workflows:doc-automation-fine-tuning |
2.53GB |
| Document Automation Indexing | - | docker pull intel/ai-workflows:doc-automation-indexing |
3.03GB |
| Fraud Detection GNN | - | docker pull intel/ai-workflows:pa-fraud-detection-gnn |
2.1GB |