# Docker Registry Paths and Example Code for Europe (Spain) (eu-south-2)
<a name="ecr-eu-south-2"></a>

The following topics list parameters for each of the algorithms and deep learning containers that are provided by Amazon SageMaker AI in this Amazon Web Services Region.

**Topics**
+ [AutoGluon (algorithm)](#autogluon-eu-south-2)
+ [BlazingText (algorithm)](#blazingtext-eu-south-2)
+ [DeepAR Forecasting (algorithm)](#forecasting-deepar-eu-south-2)
+ [Factorization Machines (algorithm)](#factorization-machines-eu-south-2)
+ [Hugging Face (algorithm)](#huggingface-eu-south-2)
+ [IP Insights (algorithm)](#ipinsights-eu-south-2)
+ [Image classification (algorithm)](#image-classification-eu-south-2)
+ [K-Means (algorithm)](#kmeans-eu-south-2)
+ [KNN (algorithm)](#knn-eu-south-2)
+ [Linear Learner (algorithm)](#linear-learner-eu-south-2)
+ [MXNet (DLC)](#mxnet-eu-south-2)
+ [Model Monitor (algorithm)](#model-monitor-eu-south-2)
+ [NTM (algorithm)](#ntm-eu-south-2)
+ [Object Detection (algorithm)](#object-detection-eu-south-2)
+ [Object2Vec (algorithm)](#object2vec-eu-south-2)
+ [PCA (algorithm)](#pca-eu-south-2)
+ [PyTorch (DLC)](#pytorch-eu-south-2)
+ [PyTorch Neuron (DLC)](#pytorch-neuron-eu-south-2)
+ [PyTorch Training Compiler (DLC)](#pytorch-training-compiler-eu-south-2)
+ [Random Cut Forest (algorithm)](#randomcutforest-eu-south-2)
+ [Scikit-learn (algorithm)](#sklearn-eu-south-2)
+ [Semantic Segmentation (algorithm)](#semantic-segmentation-eu-south-2)
+ [Seq2Seq (algorithm)](#seq2seq-eu-south-2)
+ [Spark (algorithm)](#spark-eu-south-2)
+ [Tensorflow (DLC)](#tensorflow-eu-south-2)
+ [XGBoost (algorithm)](#xgboost-eu-south-2)

## AutoGluon (algorithm)
<a name="autogluon-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='autogluon',region='eu-south-2',image_scope='inference',version='0.4')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.5.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.5.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.4.3 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.4.3 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.4.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.4.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.4.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.4.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.3.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.3.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-training:{{<tag>}} | 0.3.1 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/autogluon-inference:{{<tag>}} | 0.3.1 | inference | 

## BlazingText (algorithm)
<a name="blazingtext-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='blazingtext',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/blazingtext:{{<tag>}} | 1 | inference, training | 

## DeepAR Forecasting (algorithm)
<a name="forecasting-deepar-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='forecasting-deepar',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/forecasting-deepar:{{<tag>}} | 1 | inference, training | 

## Factorization Machines (algorithm)
<a name="factorization-machines-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='factorization-machines',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/factorization-machines:{{<tag>}} | 1 | inference, training | 

## Hugging Face (algorithm)
<a name="huggingface-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='huggingface',region='eu-south-2',version='4.4.2',image_scope='training',base_framework_version='tensorflow2.4.1')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.49.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.48.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.37.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.28.1 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.26.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.26.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.17.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.17.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.17.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.17.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.12.3 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.12.3 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.12.3 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.12.3 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.11.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.11.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.11.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.11.0 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.10.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.10.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.10.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.10.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.10.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.10.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.10.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.10.2 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.6.1 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.6.1 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.6.1 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.6.1 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-inference:{{<tag>}} | 4.6.1 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-inference:{{<tag>}} | 4.6.1 | inference | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.5.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.5.0 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-pytorch-training:{{<tag>}} | 4.4.2 | training | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/huggingface-tensorflow-training:{{<tag>}} | 4.4.2 | training | 

## IP Insights (algorithm)
<a name="ipinsights-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='ipinsights',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/ipinsights:{{<tag>}} | 1 | inference, training | 

## Image classification (algorithm)
<a name="image-classification-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='image-classification',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/image-classification:{{<tag>}} | 1 | inference, training | 

## K-Means (algorithm)
<a name="kmeans-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='kmeans',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/kmeans:{{<tag>}} | 1 | inference, training | 

## KNN (algorithm)
<a name="knn-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='knn',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/knn:{{<tag>}} | 1 | inference, training | 

## Linear Learner (algorithm)
<a name="linear-learner-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='linear-learner',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/linear-learner:{{<tag>}} | 1 | inference, training | 

## MXNet (DLC)
<a name="mxnet-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='mxnet',region='eu-south-2',version='1.4.1',py_version='py3',image_scope='inference', instance_type='ml.c5.4xlarge')
```


| Registry path | Version | Job types (image scope) | Processor types | Python versions | 
| --- | --- | --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-training:{{<tag>}} | 1.9.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference:{{<tag>}} | 1.9.0 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-training:{{<tag>}} | 1.8.0 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference:{{<tag>}} | 1.8.0 | inference | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-training:{{<tag>}} | 1.7.0 | training | CPU, GPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference:{{<tag>}} | 1.7.0 | inference | CPU, GPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference-eia:{{<tag>}} | 1.7.0 | eia | CPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-training:{{<tag>}} | 1.6.0 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference:{{<tag>}} | 1.6.0 | inference | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference-eia:{{<tag>}} | 1.5.1 | eia | CPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-training:{{<tag>}} | 1.4.1 | training | CPU, GPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference:{{<tag>}} | 1.4.1 | inference | CPU, GPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/mxnet-inference-eia:{{<tag>}} | 1.4.1 | eia | CPU | py2, py3 | 

## Model Monitor (algorithm)
<a name="model-monitor-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='model-monitor',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 437450045455.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-model-monitor-analyzer:{{<tag>}} |  | monitoring | 

## NTM (algorithm)
<a name="ntm-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='ntm',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/ntm:{{<tag>}} | 1 | inference, training | 

## Object Detection (algorithm)
<a name="object-detection-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='object-detection',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/object-detection:{{<tag>}} | 1 | inference, training | 

## Object2Vec (algorithm)
<a name="object2vec-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='object2vec',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/object2vec:{{<tag>}} | 1 | inference, training | 

## PCA (algorithm)
<a name="pca-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='pca',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/pca:{{<tag>}} | 1 | inference, training | 

## PyTorch (DLC)
<a name="pytorch-eu-south-2"></a>

For information about the supported and unsupported PyTorch versions, see the [Framework Support Policy Table](https://docs.amazonaws.cn/deep-learning-containers/latest/devguide/dlc-framework-support-policy.html) in the *Amazon Deep Learning Containers Developer Guide*.

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='pytorch',region='eu-south-2',version='1.8.0',py_version='py3',image_scope='inference', instance_type='ml.c5.4xlarge')
```


| Registry path | Version | Job types (image scope) | Processor types | Python versions | 
| --- | --- | --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.7.1 | training | CPU, GPU | py312 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.6.0 | inference | CPU, GPU | py312 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.6.0 | training | CPU, GPU | py312 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.5.1 | inference | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.5.1 | training | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.4.0 | inference | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.4.0 | inference\_graviton | CPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.4.0 | training | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.3.0 | inference | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.3.0 | inference\_graviton | CPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.3.0 | training | CPU, GPU | py311 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.2.1 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.2.0 | inference | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.2.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.1.0 | inference | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.1.0 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.1.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.0.1 | inference | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.0.1 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.0.1 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 2.0.0 | inference | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 2.0.0 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 2.0.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.13.1 | inference | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.13.1 | training | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.12.1 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-graviton:{{<tag>}} | 1.12.1 | inference\_graviton | CPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.12.1 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.12.0 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.12.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.11.0 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.11.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.10.2 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.10.2 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.10.0 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.10.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.9.1 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.9.1 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.9.0 | inference | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.9.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.8.1 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.8.1 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.8.0 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.8.0 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.7.1 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.7.1 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.6.0 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.6.0 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-eia:{{<tag>}} | 1.5.1 | eia | CPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.5.0 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.5.0 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.4.0 | inference | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.4.0 | training | CPU, GPU | py2, py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference-eia:{{<tag>}} | 1.3.1 | eia | CPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.3.1 | inference | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.3.1 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-inference:{{<tag>}} | 1.2.0 | inference | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training:{{<tag>}} | 1.2.0 | training | CPU, GPU | py2, py3 | 

## PyTorch Neuron (DLC)
<a name="pytorch-neuron-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='pytorch-neuron',region='us-west-2', image_scope='inference')
```


| Registry path | Version | Job types (image scope) | Processor types | Python versions | 
| --- | --- | --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-training-neuron:{{<tag>}} | 1.11.0 | training | TRN | py38 | 

## PyTorch Training Compiler (DLC)
<a name="pytorch-training-compiler-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='pytorch-training-compiler',region='us-west-2', version='py38')
```


| Registry path | Version | Job types (image scope) | Processor types | Python versions | 
| --- | --- | --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/pytorch-trcomp-training:{{<tag>}} | 1.12.0 | training | GPU | py38 | 

## Random Cut Forest (algorithm)
<a name="randomcutforest-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='randomcutforest',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/randomcutforest:{{<tag>}} | 1 | inference, training | 

## Scikit-learn (algorithm)
<a name="sklearn-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='sklearn',region='eu-south-2',version='0.23-1',image_scope='inference')
```


| Registry path | Version | Package version | Job types (image scope) | 
| --- | --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 1.2-1 | 1.2.1 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 1.2-1 | 1.2.1 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 1.0-1 | 1.0.2 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 1.0-1 | 1.0.2 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 1.0-1 | 1.0.2 | inference\_graviton | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 0.23-1 | 0.23.2 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 0.23-1 | 0.23.2 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 0.20.0 | 0.20.0 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-scikit-learn:{{<tag>}} | 0.20.0 | 0.20.0 | training | 

## Semantic Segmentation (algorithm)
<a name="semantic-segmentation-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='semantic-segmentation',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/semantic-segmentation:{{<tag>}} | 1 | inference, training | 

## Seq2Seq (algorithm)
<a name="seq2seq-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='seq2seq',region='eu-south-2')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/seq2seq:{{<tag>}} | 1 | inference, training | 

## Spark (algorithm)
<a name="spark-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='spark',region='eu-south-2',version='3.0',image_scope='processing')
```


| Registry path | Version | Job types (image scope) | 
| --- | --- | --- | 
| 833944533722.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-spark-processing:{{<tag>}} | 3.3 | processing | 
| 833944533722.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-spark-processing:{{<tag>}} | 3.2 | processing | 
| 833944533722.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-spark-processing:{{<tag>}} | 3.1 | processing | 
| 833944533722.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-spark-processing:{{<tag>}} | 3.0 | processing | 
| 833944533722.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-spark-processing:{{<tag>}} | 2.4 | processing | 

## Tensorflow (DLC)
<a name="tensorflow-eu-south-2"></a>

For information about the supported and unsupported TensorFlow versions, see the [Framework Support Policy Table](https://docs.amazonaws.cn/deep-learning-containers/latest/devguide/dlc-framework-support-policy.html) in the *Amazon Deep Learning Containers Developer Guide*.

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='tensorflow',region='eu-south-2',version='1.12.0',image_scope='inference',instance_type='ml.c5.4xlarge')
```


| Registry path | Version | Job types (image scope) | Processor types | Python versions | 
| --- | --- | --- | --- | --- | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.19.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.19.0 | training | CPU, GPU | py312 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.18.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.18.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.16.2 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.16.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-graviton:{{<tag>}} | 2.16.1 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.14.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-graviton:{{<tag>}} | 2.14.1 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.14.1 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.13.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-graviton:{{<tag>}} | 2.13.0 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.13.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.12.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-graviton:{{<tag>}} | 2.12.1 | inference\_graviton | CPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.12.0 | training | CPU, GPU | py310 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.11.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.11.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.11.0 | training | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.10.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.10.1 | training | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.10.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.9.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.9.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.9.2 | training | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-graviton:{{<tag>}} | 2.9.1 | inference\_graviton | CPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.8.4 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.8.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.8.0 | training | CPU, GPU | py39 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.7.1 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.7.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.6.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.6.3 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.6.2 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.6.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.6.0 | training | CPU, GPU | py38 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.5.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.5.1 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.5.0 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.4.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.4.3 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.4.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.4.1 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.3.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.3.2 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.3.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.3.1 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-eia:{{<tag>}} | 2.3.0 | eia | CPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.3.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.3.0 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.2.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.2.2 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.2.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.2.1 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.2.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.2.0 | training | CPU, GPU | py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.1.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.1.3 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.1.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.1.2 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.1.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.1.1 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.1.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.1.0 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.0.4 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.0.4 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.0.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.0.3 | training | CPU, GPU | py3, py36 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.0.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.0.2 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.0.1 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.0.1 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-eia:{{<tag>}} | 2.0.0 | eia | CPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 2.0.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 2.0.0 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.15.5 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.15.5 | training | CPU, GPU | py3, py36, py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.15.4 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.15.4 | training | CPU, GPU | py3, py36, py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.15.3 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.15.3 | training | CPU, GPU | py2, py3, py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.15.2 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.15.2 | training | CPU, GPU | py2, py3, py37 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-eia:{{<tag>}} | 1.15.0 | eia | CPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.15.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.15.0 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference-eia:{{<tag>}} | 1.14.0 | eia | CPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.14.0 | inference | CPU, GPU | - | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.14.0 | training | CPU, GPU | py2, py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-training:{{<tag>}} | 1.13.1 | training | CPU, GPU | py3 | 
| 503227376785.dkr.ecr.eu-south-2.amazonaws.com/tensorflow-inference:{{<tag>}} | 1.13.0 | inference | CPU, GPU | - | 

## XGBoost (algorithm)
<a name="xgboost-eu-south-2"></a>

The following SageMaker AI Python SDK example shows how to retrieve a specific registry path.

```
from sagemaker import image_uris
image_uris.retrieve(framework='xgboost',region='eu-south-2',version='1.5-1')
```


| Registry path | Version | Package version | Job types (image scope) | 
| --- | --- | --- | --- | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.7-1 | 1.7.4 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.7-1 | 1.7.4 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.5-1 | 1.5.2 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.5-1 | 1.5.2 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.5-1 | 1.5.2 | inference\_graviton | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.3-1 | 1.3.3 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.3-1 | 1.3.3 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.3-1 | 1.3.3 | inference\_graviton | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.2-2 | 1.2.0 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.2-2 | 1.2.0 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.2-1 | 1.2.0 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.2-1 | 1.2.0 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.0-1 | 1.0.0 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 1.0-1 | 1.0.0 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/xgboost:{{<tag>}} | 1 | 0.72 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/xgboost:{{<tag>}} | 1 | 0.72 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 0.90-2 | 0.90 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 0.90-2 | 0.90 | training | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 0.90-1 | 0.90 | inference | 
| 104374241257.dkr.ecr.eu-south-2.amazonaws.com/sagemaker-xgboost:{{<tag>}} | 0.90-1 | 0.90 | training |