Parameter Overview
You can use JobStep to create a job phase. The following is an example of defining a JobStep.
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
name |
Name of a job phase. The name contains a maximum of 64 characters, including only letters, digits, underscores (_), and hyphens (-). It must start with a letter and must be unique in a workflow. |
Yes |
str |
algorithm |
Algorithm object |
Yes |
BaseAlgorithm, Algorithm, AIGalleryAlgorithm |
spec |
Job specifications |
Yes |
JobSpec |
inputs |
Inputs of a job phase |
Yes |
JobInput or JobInput list |
outputs |
Outputs of a job phase |
Yes |
JobOutput or JobOutput list |
title |
Title for frontend display |
No |
str |
description |
Description of a job phase |
No |
str |
policy |
Phase execution policy |
No |
StepPolicy |
depend_steps |
Dependency phases |
No |
Step or step list |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
name |
Input name of the job phase. The name can contain a maximum of 64 characters, including only letters, digits, underscores (_), and hyphens (-), and must start with a letter. The input name of a step must be unique. |
Yes |
str |
data |
Input data object of a job phase |
Yes |
Dataset or OBS object. Currently, only Dataset, DatasetPlaceholder, DatasetConsumption, OBSPath, OBSConsumption, OBSPlaceholder, and DataConsumptionSelector are supported. |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
name |
Output name of the job phase. The name can contain a maximum of 64 characters, including only letters, digits, underscores (_), and hyphens (-), and must start with a letter. The output name of a step must be unique. |
Yes |
str |
obs_config |
OBS output configuration |
No |
OBSOutputConfig |
model_config |
Model output configuration |
No |
ModelConfig |
metrics_config |
Metrics configuration |
No |
MetricsConfig |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
obs_path |
Existing OBS directory |
Yes |
str, Placeholder, Storage |
metric_file |
Name of the file that stores metric information |
No |
str, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
id |
Algorithm ID |
No |
str |
subscription_id |
Subscription ID of the subscribed algorithm |
No |
str |
item_version_id |
Version ID of the subscribed algorithm |
No |
str |
code_dir |
Code directory |
No |
str, Placeholder, Storage |
boot_file |
Boot file |
No |
str, Placeholder, Storage |
command |
Boot command |
No |
str, Placeholder |
parameters |
Algorithm hyperparameters |
No |
AlgorithmParameters list |
engine |
Information about the image used by the job |
No |
JobEngine |
environments |
Environment variables |
No |
dict |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
algorithm_id |
Algorithm ID |
Yes |
str |
parameters |
Algorithm hyperparameters |
No |
AlgorithmParameters list |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
subscription_id |
Subscription ID of the subscribed algorithm |
Yes |
str |
item_version_id |
Version ID of the subscribed algorithm |
Yes |
str |
parameters |
Algorithm hyperparameters |
No |
AlgorithmParameters list |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
name |
Name of an algorithm hyperparameter |
Yes |
str |
value |
Value of an algorithm hyperparameter |
Yes |
int, bool, float, str, Placeholder, Storage |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
engine_id |
Image ID |
No |
str, Placeholder |
engine_name |
Image name |
No |
str, Placeholder |
engine_version |
Image version |
No |
str, Placeholder |
image_url |
Image URL |
No |
str, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
resource |
Resource |
Yes |
JobResource |
log_export_path |
Log output path |
No |
LogExportPath |
schedule_policy |
Job scheduling policy |
No |
SchedulePolicy |
volumes |
Information about the file system mounted to the job |
No |
list[Volume] |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
flavor |
Resource specifications |
Yes |
Placeholder |
node_count |
Number of nodes. The default value is 1. If there are multiple nodes, distributed training is supported. |
No |
int, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
priority |
Job scheduling priority. The value can only be 1, 2, or 3, indicating low, medium, and high priorities, respectively. |
Yes |
int, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
nfs |
NFS file system object |
No |
NFS |
pfs |
OBS parallel file system object. In a volume object, either PFS or NFS can be specified. |
No |
PFS, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
nfs_server_path |
Service address of the NFS file system. |
Yes |
str, Placeholder |
local_path |
Path mounted to the container. |
Yes |
str, Placeholder |
read_only |
Indicates if the mount mode is set to read-only. |
No |
bool, Placeholder |
Parameter |
Description |
Mandatory |
Data Type |
---|---|---|---|
pfs_path |
Path of the parallel file system |
Yes |
str, Placeholder |
local_path |
Path mounted to the container. |
Yes |
str, Placeholder |
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot