本文属于机器翻译版本。若本译文内容与英语原文存在差异，则一律以英文原文为准。
非洲（开普敦）Docker Regarning 路径和示例代码
以下主题列出了该区域中由 Amazon 提供的每种算法和深度学习容器的参数 SageMaker.
主题
- AutoGluon （算法）
- BlazingText （算法）
- Chainer (DLC)
- 澄清（算法）
- Data Wrangler（算法）
- 调试器（算法）
- DeepAR 预测（算法）
- 分解机（算法）
- 拥抱脸（算法）
- IP 洞察（算法）
- 图像分类（算法）
- Inferentia mxNet (DLC)
- Inferentia PyTorch (DLC)
- K-均值（算法）
- KNN（算法）
- 线性学习器（算法）
- mxNet (DLC)
- mxNet Coach (DLC)
- 模型监视器（算法）
- NTM（算法）
- Neo 图像分类（算法）
- Neo mxNet (DLC)
- 新 PyTorch (DLC)
- Neo Tensorflow (DLC)
- Neo xgBoost（算法）
- 物体检测（算法）
- Object2Vec（算法）
- PCA（算法）
- PyTorch (DLC)
- 随机砍伐森林（算法）
- Scikit-Learn（算法）
- 语义分割（算法）
- Seq2Seq（算法）
- 火花（算法）
- SparkML 服务（算法）
- Tensorflow (DLC)
- Tensorflow 教练 (DLC)
- Tensorflow 推论 (DLC)
- 张量流射线 (DLC)
- xgBoost（算法）
AutoGluon （算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='autogluon',region='af-south-1',image_scope='inference',version='0.4')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.5.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.5.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.4.3 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.4.3 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.4.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.4.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.4.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.4.0 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.3.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.3.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-training：<tag> |
0.3.1 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/autogluon-inference：<tag> |
0.3.1 | 推断 |
BlazingText （算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='blazingtext',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/blazingtext：<tag> |
1 | 推理 |
Chainer (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='chainer',region='af-south-1',version='5.0.0',py_version='py3',image_scope='inference',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-chainer：<tag> |
5.0.0 | 推理 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-chainer：<tag> |
4.1.0 | 推理 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-chainer：<tag> |
4.0.0 | 推理 | CPU、GPU | py2、py3 |
澄清（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='clarify',region='af-south-1',version='1.0',image_scope='processing')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
811711786498.dkr.ecr.af-south-1.amazonaws.com/sagemaker-clarify-processing：<tag> |
1.0 | 处理 |
Data Wrangler（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='data-wrangler',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
143210264188.dkr.ecr.af-south-1.amazonaws.com/sagemaker-data-wrangler-container：<tag> |
1.x | 处理 |
调试器（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='debugger',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
314341159256.dkr.ecr.af-south-1.amazonaws.com/sagemaker-debugger-rules：<tag> |
latest | 调试器 |
DeepAR 预测（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='forecasting-deepar',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/forecasting-deepar：<tag> |
1 | 推理，培训 |
分解机（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='factorization-machines',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/factorization-machines：<tag> |
1 | 推理，培训 |
拥抱脸（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='huggingface',region='af-south-1',version='4.4.2',image_scope='training',base_framework_version='tensorflow2.4.1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.17.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.17.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.17.0 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.1.0 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.12.3 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.12.3 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.12.3 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.12.3 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.11.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.11.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.11.0 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.11.0 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.10.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.10.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.10.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.10.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.10.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.10.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.10.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.10.2 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.6.1 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.6.1 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.6.1 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.6.1 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-inference：<tag> |
4.6.1 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-inference：<tag> |
4.6.1 | 推断 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.5.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.5.0 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-pytorch-training：<tag> |
4.4.2 | 培训 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/huggingface-tensorflow-training：<tag> |
4.4.2 | 培训 |
IP 洞察（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='ipinsights',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/ipinsights：<tag> |
1 | 推理 |
图像分类（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='image-classification',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/image-classification：<tag> |
1 | 推理 |
Inferentia mxNet (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='inferentia-mxnet',region='af-south-1',version='1.5.1',instance_type='ml.inf1.6xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-mxnet：<tag> |
1.8 | 推断 | inf | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-mxnet：<tag> |
1.5.1 | 推断 | inf | py3.3 |
Inferentia PyTorch (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='inferentia-pytorch',region='af-south-1',version='1.9',py_version='py3')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-pytorch：<tag> |
1.9 | 推断 | inf | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-pytorch：<tag> |
1.8 | 推断 | inf | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-pytorch：<tag> |
1.7 | 推断 | inf | py3.3 |
K-均值（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='kmeans',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/kmeans：<tag> |
1 | 推理 |
KNN（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='knn',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/knn：<tag> |
1 | 推理 |
线性学习器（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='linear-learner',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/linear-learner：<tag> |
1 | 推理 |
mxNet (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='mxnet',region='af-south-1',version='1.4.1',py_version='py3',image_scope='inference', instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-training：<tag> |
1.9.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference：<tag> |
1.9.0 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-training：<tag> |
1.8.0 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference：<tag> |
1.8.0 | 推断 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-training：<tag> |
1.7.0 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference：<tag> |
1.7.0 | 推断 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference-eia：<tag> |
1.7.0 | EIA | CPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-training：<tag> |
1.6.0 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference：<tag> |
1.6.0 | 推断 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference-eia：<tag> |
1.5.1 | EIA | CPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.4.1 | 培训 | CPU、GPU | py2 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-training：<tag> |
1.4.1 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet-serving：<tag> |
1.4.1 | 推断 | CPU、GPU | py2 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference：<tag> |
1.4.1 | 推断 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/mxnet-inference-eia：<tag> |
1.4.1 | EIA | CPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.4.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet-serving：<tag> |
1.4.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet-serving-eia：<tag> |
1.4.0 | EIA | CPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.3.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.3.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet-eia：<tag> |
1.3.0 | EIA | CPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.2.1 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.2.1 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.1.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.1.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.0.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
1.0.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
0.12.1 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-mxnet：<tag> |
0.12.1 | 推断 | CPU、GPU | py2、py3 |
mxNet Coach (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='coach-mxnet',region='af-south-1',version='0.11',py_version='py3',image_scope='training',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-mxnet:coach0.11.0-<tag> |
0.11.0 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-mxnet:coach0.11-<tag> |
0.11 | 培训 | CPU、GPU | py3.3 |
模型监视器（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='model-monitor',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
875698925577.dkr.ecr.af-south-1.amazonaws.com/sagemaker-model-monitor-analyzer：<tag> |
监控 |
NTM（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='ntm',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/ntm：<tag> |
1 | 推理 |
Neo 图像分类（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='image-classification-neo',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/image-classification-neo：<tag> |
latest | 推断 |
Neo mxNet (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='neo-mxnet',region='af-south-1',version='1.8',py_version='py3',image_scope='inference', instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-mxnet：<tag> |
1.8 | 推断 | CPU、GPU | py3.3 |
新 PyTorch (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='neo-pytorch',region='af-south-1',version='1.6',image_scope='inference',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-pytorch：<tag> |
1.8 | 推断 | CPU、GPU | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-pytorch：<tag> |
1.7 | 推断 | CPU、GPU | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-pytorch：<tag> |
1.6 | 推断 | CPU、GPU | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-pytorch：<tag> |
1.5 | 推断 | CPU、GPU | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-pytorch：<tag> |
1.4 | 推断 | CPU、GPU | py3.3 |
Neo Tensorflow (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='neo-tensorflow',region='af-south-1',version='1.15.3',py_version='py3',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-tensorflow：<tag> |
2.4.2 | 推断 | CPU、GPU | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-inference-tensorflow：<tag> |
1.15.3 | 推断 | CPU、GPU | py3.3 |
Neo xgBoost（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='xgboost-neo',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/xgboost-neo：<tag> |
latest | 推断 |
物体检测（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='object-detection',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/object-detection：<tag> |
1 | 推理 |
Object2Vec（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='object2vec',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/object2vec：<tag> |
1 | 推理 |
PCA（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='pca',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/pca：<tag> |
1 | 推理 |
PyTorch (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='pytorch',region='af-south-1',version='1.8.0',py_version='py3',image_scope='inference', instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.12.0 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.12.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.11.0 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.11.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.10.2 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.10.2 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.10.0 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.10.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.9.1 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.9.1 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.9.0 | 推断 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.9.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.8.1 | 推断 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.8.1 | 培训 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.8.0 | 推断 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.8.0 | 培训 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.7.1 | 推断 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.7.1 | 培训 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.6.0 | 推断 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.6.0 | 培训 | CPU、GPU | py3、py36 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.5.0 | 推断 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.5.0 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.4.0 | 推断 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.4.0 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.3.1 | 推断 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.3.1 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-inference：<tag> |
1.2.0 | 推断 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/pytorch-training：<tag> |
1.2.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
1.1.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
1.1.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
1.0.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
1.0.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
0.4.0 | 推断 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-pytorch：<tag> |
0.4.0 | 培训 | CPU、GPU | py2、py3 |
随机砍伐森林（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='randomcutforest',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/randomcutforest：<tag> |
1 | 推理 |
Scikit-Learn（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='sklearn',region='af-south-1',version='0.23-1',image_scope='inference')
| 注册表路径 | 版本 | 程序包版本 | Job 类型（图像范围） |
|---|---|---|---|
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-scikit-learn：<tag> |
1.0-1 | 1.0.2 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-scikit-learn：<tag> |
0.23-1 | 0.23.2 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-scikit-learn：<tag> |
0.20.0 | 0.20.0 | 推理，培训 |
语义分割（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='semantic-segmentation',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/semantic-segmentation：<tag> |
1 | 推理，培训 |
Seq2Seq（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='seq2seq',region='af-south-1')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
455444449433.dkr.ecr.af-south-1.amazonaws.com/seq2seq：<tag> |
1 | 推理，培训 |
火花（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='spark',region='af-south-1',version='3.0',image_scope='processing')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
309385258863.dkr.ecr.af-south-1.amazonaws.com/sagemaker-spark-processing：<tag> |
3.1 | 处理 |
309385258863.dkr.ecr.af-south-1.amazonaws.com/sagemaker-spark-processing：<tag> |
3.0 | 处理 |
309385258863.dkr.ecr.af-south-1.amazonaws.com/sagemaker-spark-processing：<tag> |
2.4 | 处理 |
SparkML 服务（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='sparkml-serving',region='af-south-1',version='2.4')
| 注册表路径 | 版本 | Job 类型（图像范围） |
|---|---|---|
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-sparkml-serving：<tag> |
2.4 | 推断 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-sparkml-serving：<tag> |
2.2 | 推断 |
Tensorflow (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='tensorflow',region='af-south-1',version='1.12.0',image_scope='inference',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.9.1 | 培训 | CPU、GPU | py39 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.8.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.8.0 | 培训 | CPU、GPU | py39 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.7.1 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.7.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.6.3 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.6.3 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.6.2 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.6.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.6.0 | 培训 | CPU、GPU | py38 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.5.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.5.1 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.5.0 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.4.3 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.4.3 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.4.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.4.1 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.3.2 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.3.2 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.3.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.3.1 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference-eia：<tag> |
2.3.0 | EIA | CPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.3.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.3.0 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.2.2 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.2.2 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.2.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.2.1 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.2.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.2.0 | 培训 | CPU、GPU | py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.1.3 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.1.3 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.1.2 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.1.2 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.1.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.1.1 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.1.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.1.0 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.0.4 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.0.4 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.0.3 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.0.3 | 培训 | CPU、GPU | py3.3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.0.2 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.0.2 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.0.1 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.0.1 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference-eia：<tag> |
2.0.0 | EIA | CPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
2.0.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
2.0.0 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.15.5 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.15.5 | 培训 | CPU、GPU | py3、py36、py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.15.4 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.15.4 | 培训 | CPU、GPU | py3、py36、py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.15.3 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.15.3 | 培训 | CPU、GPU | py2、py3、py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.15.2 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.15.2 | 培训 | CPU、GPU | py2、py3、py37 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference-eia：<tag> |
1.15.0 | EIA | CPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.15.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.15.0 | 培训 | CPU、GPU | py2、py3 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference-eia：<tag> |
1.14.0 | EIA | CPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.14.0 | 推断 | CPU、GPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.14.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-scriptmode：<tag> |
1.13.1 | 培训 | CPU、GPU | py2 |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-training：<tag> |
1.13.1 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-serving-eia：<tag> |
1.13.0 | EIA | CPU | - |
626614931356.dkr.ecr.af-south-1.amazonaws.com/tensorflow-inference：<tag> |
1.13.0 | 推断 | CPU、GPU | - |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-serving-eia：<tag> |
1.12.0 | EIA | CPU | - |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-serving：<tag> |
1.12.0 | 推断 | CPU、GPU | - |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-scriptmode：<tag> |
1.12.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-serving-eia：<tag> |
1.11.0 | EIA | CPU | - |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-serving：<tag> |
1.11.0 | 推断 | CPU、GPU | - |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-scriptmode：<tag> |
1.11.0 | 培训 | CPU、GPU | py2、py3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow-eia：<tag> |
1.10.0 | EIA | CPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.10.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.10.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.9.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.9.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.8.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.8.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.7.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.7.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.6.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.6.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.5.0 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.5.0 | 培训 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.4.1 | 推断 | CPU、GPU | py2 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-tensorflow：<tag> |
1.4.1 | 培训 | CPU、GPU | py2 |
Tensorflow 教练 (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='coach-tensorflow',region='af-south-1',version='1.0.0',image_scope='training',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:coach0.11.1-<tag> |
0.11.1 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:coach0.11.0-<tag> |
0.11.0 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:coach0.11-<tag> |
0.11 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:coach0.10.1-<tag> |
0.10.1 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:coach0.10-<tag> |
0.10 | 培训 | CPU、GPU | py3.3 |
Tensorflow 推论 (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='inferentia-tensorflow',region='af-south-1',version='1.15.0',instance_type='ml.inf1.6xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-tensorflow：<tag> |
2.5.2 | 推断 | inf | py3.3 |
774647643957.dkr.ecr.af-south-1.amazonaws.com/sagemaker-neo-tensorflow：<tag> |
1.15.0 | 推断 | inf | py3.3 |
张量流射线 (DLC)
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='ray-tensorflow',region='af-south-1',version='0.8.5',instance_type='ml.c5.4xlarge')
| 注册表路径 | 版本 | Job 类型（图像范围） | 处理器类型 | Python 版本 |
|---|---|---|---|---|
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:ray0.6.5-<tag> |
0.6.5 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:ray0.6-<tag> |
0.6 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:ray0.5.3-<tag> |
0.5.3 | 培训 | CPU、GPU | py3.3 |
313743910680.dkr.ecr.af-south-1.amazonaws.com/sagemaker-rl-tensorflow:ray0.5-<tag> |
0.5 | 培训 | CPU、GPU | py3.3 |
xgBoost（算法）
SageMaker 检索注册表路径的 Python SDK 示例。
from sagemaker import image_uris image_uris.retrieve(framework='xgboost',region='af-south-1',version='1.5-1')
| 注册表路径 | 版本 | 程序包版本 | Job 类型（图像范围） |
|---|---|---|---|
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
1.5-1 | 1.5.2 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
1.3.1 | 1.3.3 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
1.2-2 | 1.2.0 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
1.2-1 | 1.2.0 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
1.0-1 | 1.0 | 推理，培训 |
455444449433.dkr.ecr.af-south-1.amazonaws.com/xgboost：<tag> |
1 | 0.72 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
0.90-2 | 0.90 | 推理，培训 |
510948584623.dkr.ecr.af-south-1.amazonaws.com/sagemaker-xgboost：<tag> |
0.90-1 | 0.90 | 推理，培训 |