Skip to content

Commit 6744e0d

Browse files
laurenyulaurenyu
authored
doc: update container dependency versions for MXNet and PyTorch (#1233)
1 parent 6d41d24 commit 6744e0d

File tree

2 files changed

+60
-54
lines changed

2 files changed

+60
-54
lines changed

‎src/sagemaker/mxnet/README.rst‎

Lines changed: 13 additions & 13 deletions
Original file line numberDiff line numberDiff line change
@@ -21,19 +21,19 @@ Your MXNet training script will be run on version 1.2.1 by default. (See below f
2121

2222
The Docker images have the following dependencies installed:
2323

24-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
25-
| Dependencies | MXNet 0.12.1 | MXNet 1.0.0 | MXNet 1.1.0 | MXNet 1.2.1 | MXNet 1.3.0 | MXNet 1.4.0 | MXNet 1.4.1 |
26-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
27-
| Python | 2.7 or 3.5 | 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.6| 2.7 or 3.6|
28-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
29-
| CUDA (GPU image only) | 9.0 | 9.0 | 9.0 | 9.0 | 9.0 | 9.2 | 10.0 |
30-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
31-
| numpy | 1.13.3 | 1.13.3 | 1.13.3 | 1.14.5 | 1.14.6 | 1.16.3 | 1.14.5 |
32-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
33-
| onnx | N/A | N/A | N/A | 1.2.1 | 1.2.1 | 1.4.1 | 1.4.1 |
34-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
35-
| keras-mxnet | N/A | N/A | N/A | N/A | 2.2.2 | 2.2.4.1 | 2.2.4.1 |
36-
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+
24+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
25+
| Dependencies | MXNet 0.12.1 | MXNet 1.0.0 | MXNet 1.1.0 | MXNet 1.2.1 | MXNet 1.3.0 | MXNet 1.4.0 | MXNet 1.4.1 | MXNet 1.6.0 |
26+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
27+
| Python | 2.7 or 3.5 | 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.5| 2.7 or 3.6| 2.7 or 3.6| 2.7 or 3.6|
28+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
29+
| CUDA (GPU image only) | 9.0 | 9.0 | 9.0 | 9.0 | 9.0 | 9.2 | 10.0 | 10.1 |
30+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
31+
| numpy | 1.13.3 | 1.13.3 | 1.13.3 | 1.14.5 | 1.14.6 | 1.16.3 | 1.14.5 | 1.17.2 |
32+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
33+
| onnx | N/A | N/A | N/A | 1.2.1 | 1.2.1 | 1.4.1 | 1.4.1 | 1.6.0 |
34+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
35+
| keras-mxnet | N/A | N/A | N/A | N/A | 2.2.2 | 2.2.4.1 | 2.2.4.1 | 2.2.4.2 |
36+
+-------------------------+--------------+-------------+-------------+-------------+-------------+-------------+-------------+-------------+
3737

3838
The Docker images extend Ubuntu 16.04.
3939

‎src/sagemaker/pytorch/README.rst‎

Lines changed: 47 additions & 41 deletions
Original file line numberDiff line numberDiff line change
@@ -31,59 +31,65 @@ libraries installed. When creating the Estimator and calling deploy to create th
3131
the environment your script runs in.
3232

3333
SageMaker runs PyTorch Estimator scripts in either Python 2 or Python 3. You can select the Python version by
34-
passing a ``py_version`` keyword arg to the PyTorch Estimator constructor. Setting this to `py3` (the default) will cause your
35-
training script to be run on Python 3.5. Setting this to `py2` will cause your training script to be run on Python 2.7
34+
passing a ``py_version`` keyword arg to the PyTorch Estimator constructor. Setting this to ``py3`` (the default) will cause your
35+
training script to be run on Python 3.5. Setting this to ``py2`` will cause your training script to be run on Python 2.7
3636
This Python version applies to both the Training Job, created by fit, and the Endpoint, created by deploy.
3737

3838
The PyTorch Docker images have the following dependencies installed:
3939

40-
+-----------------------------+---------------+-------------------+-------------------+
41-
| Dependencies | pytorch 0.4.0 | pytorch 1.0.0 | pytorch 1.1.0 |
42-
+-----------------------------+---------------+-------------------+-------------------+
43-
| boto3 | >=1.7.35 | >=1.9.11 | 1.9.82 |
44-
+-----------------------------+---------------+-------------------+-------------------+
45-
| botocore | >=1.10.35 | >=1.12.11 | >= 1.12.11 |
46-
+-----------------------------+---------------+-------------------+-------------------+
47-
| CUDA (GPU image only) | 9.0 | 9.0 | 10.1 |
48-
+-----------------------------+---------------+-------------------+-------------------+
49-
| numpy | >=1.14.3 | >=1.15.2 | 1.16.4 |
50-
+-----------------------------+---------------+-------------------+-------------------+
51-
| Pillow | >=5.1.0 | >=5.2.0 | 6.0.0 |
52-
+-----------------------------+---------------+-------------------+-------------------+
53-
| pip | >=10.0.1 | >=18.0 | >=18.0 |
54-
+-----------------------------+---------------+-------------------+-------------------+
55-
| python-dateutil | >=2.7.3 | >=2.7.3 | >=2.7.3 |
56-
+-----------------------------+---------------+-------------------+-------------------+
57-
| retrying | >=1.3.3 | >=1.3.3 | 1.3.3 |
58-
+-----------------------------+---------------+-------------------+-------------------+
59-
| s3transfer | >=0.1.13 | >=0.1.13 | >=0.1.13 |
60-
+-----------------------------+---------------+-------------------+-------------------+
61-
| sagemaker-containers | >=2.1.0 | >=2.1.0 | 2.4.10.post0 |
62-
+-----------------------------+---------------+-------------------+-------------------+
63-
| sagemaker-pytorch-container | 1.0 | 1.1 | 1.2 |
64-
+-----------------------------+---------------+-------------------+-------------------+
65-
| setuptools | >=39.2.0 | >=40.4.3 | >=40.4.3 |
66-
+-----------------------------+---------------+-------------------+-------------------+
67-
| six | >=1.11.0 | >=1.11.0 | 1.12.0 |
68-
+-----------------------------+---------------+-------------------+-------------------+
69-
| torch | 0.4.0 | 1.0.0 | 1.1.0 |
70-
+-----------------------------+---------------+-------------------+-------------------+
71-
| torchvision | 0.2.1 | 0.2.1 | 0.3.0 |
72-
+-----------------------------+---------------+-------------------+-------------------+
73-
| Python | 2.7 or 3.5 | 2.7 or 3.6 | 2.7 or 3.6 |
74-
+-----------------------------+---------------+-------------------+-------------------+
40+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
41+
| Dependencies | pytorch 0.4.0 | pytorch 1.0.0 | pytorch 1.1.0 | pytorch 1.2.0 | pytorch 1.3.1 |
42+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
43+
| boto3 | >=1.7.35 | >=1.9.11 | 1.9.82 | 1.9.249 | 1.10.34 |
44+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
45+
| botocore | >=1.10.35 | >=1.12.11 | >= 1.12.11 | 1.12.249 | 1.13.34 |
46+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
47+
| CUDA (GPU image only) | 9.0 | 9.0 | 10.1 | 10.0 | 10.1 |
48+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
49+
| numpy | >=1.14.3 | >=1.15.2 | 1.16.4 | 1.16.4 | 1.16.4 |
50+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
51+
| Pillow | >=5.1.0 | >=5.2.0 | 6.0.0 | 5.4.1 | 6.2.1 |
52+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
53+
| pip | >=10.0.1 | >=18.0 | >=18.0 | 19.3 | 19.3.1 |
54+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
55+
| python-dateutil | >=2.7.3 | >=2.7.3 | >=2.7.3 | 2.8.0 | 2.8.0 |
56+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
57+
| retrying | >=1.3.3 | >=1.3.3 | 1.3.3 | 1.3.3 | 1.3.3 |
58+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
59+
| s3transfer | >=0.1.13 | >=0.1.13 | >=0.1.13 | 0.2.1 | 0.2.1 |
60+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
61+
| sagemaker-containers | >=2.1.0 | >=2.1.0 | 2.4.10.post0 | 2.5.4 | 2.5.4 |
62+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
63+
| sagemaker-inference (inference only) | N/A | N/A | N/A | N/A | 1.1.2 |
64+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
65+
| sagemaker-pytorch-container | 1.0 | 1.1 | 1.2 | 1.2 | 1.3 |
66+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
67+
| setuptools | >=39.2.0 | >=40.4.3 | >=40.4.3 | 41.4.0 | 42.0.2.post20191203 |
68+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
69+
| six | >=1.11.0 | >=1.11.0 | 1.12.0 | 1.12.0 | 1.12.0 |
70+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
71+
| torch | 0.4.0 | 1.0.0 | 1.1.0 | 1.2.0 | 1.3.1 |
72+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
73+
| torchvision | 0.2.1 | 0.2.1 | 0.3.0 | 0.4.0a0+9232c4a | 0.4.2 |
74+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
75+
| Python | 2.7 or 3.5 | 2.7 or 3.6 | 2.7 or 3.6 | 2.7 or 3.6 | 2.7 or 3.6 |
76+
+---------------------------------------+---------------+----------------+---------------+-----------------+---------------------+
7577

7678
The Docker images extend Ubuntu 16.04.
7779

78-
If you need to install other dependencies you can put them into `requirements.txt` file and put it in the source directory
80+
With most versions of PyTorch, if you need to install other dependencies, you can put them into ``requirements.txt`` file and put it in the source directory
7981
(``source_dir``) you provide to the `PyTorch Estimator <#pytorch-estimators>`__.
82+
For more, including directions specific to each version of PyTorch, see `Using third-party libraries <https://sagemaker.readthedocs.io/en/stable/using_pytorch.html#using-third-party-libraries>`_.
8083

8184
You can select version of PyTorch by passing a ``framework_version`` keyword arg to the PyTorch Estimator constructor.
8285
Currently supported versions are listed in the above table. You can also set ``framework_version`` to only specify major and
8386
minor version, which will cause your training script to be run on the latest supported patch version of that minor
8487
version.
8588

86-
Alternatively, you can build your own image by following the instructions in the SageMaker Chainer containers
87-
repository, and passing ``image_name`` to the Chainer Estimator constructor.
89+
Alternatively, you can build your own image by following the instructions in the SageMaker PyTorch containers
90+
repository, and passing ``image_name`` to the PyTorch Estimator constructor.
8891

89-
You can visit `the SageMaker PyTorch containers repository <https://github.com/aws/sagemaker-pytorch-containers>`_.
92+
You can visit the SageMaker PyTorch containers repositories:
93+
94+
- training: https://github.com/aws/sagemaker-pytorch-container
95+
- serving: https://github.com/aws/sagemaker-pytorch-serving-container

0 commit comments

Comments
 (0)