From eebf6cb7fac542582fedd871f1d64a381d2a54eb Mon Sep 17 00:00:00 2001 From: "Lai, Weijian" Date: Thu, 11 May 2023 09:58:56 +0000 Subject: [PATCH] modelarts_sdk-ref_20230504 Reviewed-by: Jiang, Beibei Co-authored-by: Lai, Weijian Co-committed-by: Lai, Weijian --- docs/modelarts/sdk-ref/modelarts_04_0131.html | 172 +++++++++--------- 1 file changed, 86 insertions(+), 86 deletions(-) diff --git a/docs/modelarts/sdk-ref/modelarts_04_0131.html b/docs/modelarts/sdk-ref/modelarts_04_0131.html index 2aec2e29..543796c0 100644 --- a/docs/modelarts/sdk-ref/modelarts_04_0131.html +++ b/docs/modelarts/sdk-ref/modelarts_04_0131.html @@ -135,213 +135,213 @@

Parameter Description

-
Table 1 Estimator request parameters

Parameter

+
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Table 1 Estimator request parameters

Parameter

Mandatory

+

Mandatory

Type

+

Type

Description

+

Description

modelarts_session

+

modelarts_session

Yes

+

Yes

Object

+

Object

Session object. For details about the initialization method, see Session Authentication.

+

Session object. For details about the initialization method, see Session Authentication.

train_instance_count

+

train_instance_count

Yes

+

Yes

Long

+

Long

Number of compute nodes in a training job

+

Number of compute nodes in a training job

code_dir

+

code_dir

No

+

No

String

+

String

Code directory of a training job, for example, /bucket/src/. Leave this parameter blank when model_name is set.

+

Code directory of a training job, for example, /bucket/src/. Leave this parameter blank when model_name is set.

boot_file

+

boot_file

No

+

No

String

+

String

Boot file of a training job, which needs to be stored in the code directory. For example, /bucket/src/boot.py. Leave this parameter blank when model_name is set.

+

Boot file of a training job, which needs to be stored in the code directory. For example, /bucket/src/boot.py. Leave this parameter blank when model_name is set.

output_path

+

output_path

Yes

+

Yes

String

+

String

Output path of a training job

+

Output path of a training job

hyperparameters

+

hyperparameters

No

+

No

JSON Array

+

JSON Array

Running parameters of a training job. It is a collection of label-value pairs of the string type. This parameter is a container environment variable when a job uses a custom image. For details about hyperparameters if a built-in algorithm is used, see Algorithms and Their Running Parameters.

+

Running parameters of a training job. It is a collection of label-value pairs of the string type. This parameter is a container environment variable when a job uses a custom image. For details about hyperparameters if a built-in algorithm is used, see Algorithms and Their Running Parameters.

log_url

+

log_url

No

+

No

String

+

String

OBS URL of the logs of a training job. By default, this parameter is left blank. Example value: /usr/log/

+

OBS URL of the logs of a training job. By default, this parameter is left blank. Example value: /usr/log/

train_instance_type

+

train_instance_type

Yes

+

Yes

Long

+

Long

Resource flavor selected for a training job. If you choose to train on the training platform, obtain the value by calling the API described in Obtaining Resource Flavors.

+

Resource flavor selected for a training job. If you choose to train on the training platform, obtain the value by calling the API described in Obtaining Resource Flavors.

framework_type

+

framework_type

No

+

No

String

+

String

Engine selected for a training job. Obtain the value by calling the API described in Obtaining Engine Types. Leave this parameter blank when model_name is set.

+

Engine selected for a training job. Obtain the value by calling the API described in Obtaining Engine Types. Leave this parameter blank when model_name is set.

framework_version

+

framework_version

No

+

No

String

+

String

Engine version selected for a training job. Obtain the value by calling the API described in Obtaining Engine Types. Leave this parameter blank when model_name is set.

+

Engine version selected for a training job. Obtain the value by calling the API described in Obtaining Engine Types. Leave this parameter blank when model_name is set.

job_description

+

job_description

No

+

No

String

+

String

Description of a training job

+

Description of a training job

user_image_url

+

user_image_url

No

+

No

String

+

String

SWR URL of the custom image used by a training job. Example value: 100.125.5.235:20202/jobmng/custom-cpu-base:1.0

+

SWR URL of the custom image used by a training job. Example value: 100.125.5.235:20202/jobmng/custom-cpu-base:1.0

user_command

+

user_command

No

+

No

String

+

String

Boot command used to start the container of the custom image of a training job. The format is bash /home/work/run_train.sh python /home/work/user-job-dir/app/train.py {python_file_parameter}.

+

Boot command used to start the container of the custom image of a training job. The format is bash /home/work/run_train.sh python /home/work/user-job-dir/app/train.py {python_file_parameter}.
-
Table 2 fit request parameters

Parameter

+
- - - - - - - - - - - - - - - - - - - - - - -
Table 2 fit request parameters

Parameter

Mandatory

+

Mandatory

Type

+

Type

Description

+

Description

inputs

+

inputs

Yes

+

Yes

String

+

String

Data storage location of a training job.

+

Data storage location of a training job.

inputs cannot be used with dataset_id and dataset_version_id, or with data_source at the same time. However, one of the parameters must exist.

Only this parameter is supported in local training.

dataset_id

+

dataset_id

No

+

No

String

+

String

Dataset ID of a training job. To obtain the dataset ID, view basic information about the dataset.

+

Dataset ID of a training job. To obtain the dataset ID, Managing Dataset Versions.

This parameter must be used together with dataset_version_id, but cannot be used together with inputs.

dataset_version_id

+

dataset_version_id

No

+

No

String

+

String

Dataset version ID of a training job. To obtain the dataset version ID, view basic information about the dataset.

+

Dataset version ID of a training job. To obtain the dataset version ID, Managing Dataset Versions.

This parameter must be used together with dataset_id, but cannot be used together with inputs.

wait

+

wait

No

+

No

Boolean

+

Boolean

Whether to wait for the completion of a training job. Default value: False

+

Whether to wait for the completion of a training job. Default value: False

job_name

+

job_name

No

+

No

String

+

String

Name of a training job, consisting of 1 to 64 alphanumeric characters. If this parameter is left blank, a job name is generated randomly.

+

Name of a training job, consisting of 1 to 64 alphanumeric characters. If this parameter is left blank, a job name is generated randomly.
-
Table 3 Parameters in the successful response to training

Parameter

+
- - - - -
Table 3 Parameters in the successful response to training

Parameter

Type

+

Type

Description

+

Description

TrainingJob

+

TrainingJob

Object

+

Object

Training object. This object contains attributes such as job_id and version_id, and operations on a training job, such as querying, modifying, or deleting the training job. For example, you can use job_instance.job_id to obtain the ID of a training job.

+

Training object. This object contains attributes such as job_id and version_id, and operations on a training job, such as querying, modifying, or deleting the training job. For example, you can use job_instance.job_id to obtain the ID of a training job.