You can use the Deep Learning Containers (DLC) client to submit various types of training jobs. This topic describes the submission commands, including the call format, parameter descriptions, and examples.
Common parameters for submitting jobs
The following common parameters apply when you use the DLC command line interface to submit TensorFlow (tfjob), PyTorch (pytorchjob), or XGBoost (xgboostjob) jobs.
Table 1. Common parameters for submitting jobs
Parameter | Required | Description | Type | Supported in job parameter description file |
name | Yes | The name of the job. Multiple jobs can have the same name. | STRING | Yes |
command | Yes | The start command for each node. | STRING | Yes |
data_sources | No | The IDs of the attached datasets. Go to the Dataset page to view the IDs. Separate multiple IDs with commas (,). The default value is empty. | STRING | Yes |
code_source | No | The ID of the code build. Go to the Code Configuration page to view the ID. Only a single ID is supported. The default value is empty. | STRING | Yes |
code_branch | No | The branch of the code repository. Use this parameter with the code_source parameter. | STRING | Yes |
code_commit | No | The commit ID of the code repository. Use this parameter with the code_source parameter. | STRING | Yes |
thirdparty_libs | No | The third-party Python libraries. Separate multiple libraries with commas (,). The default value is empty. | STRING | Yes |
thirdparty_lib_dir | No | The directory that contains the requirements.txt file for installing third-party Python libraries. The default value is empty. | STRING | No |
vpc_id | No | The ID of the virtual private cloud (VPC) that the job can access. The default value is empty. | STRING | Yes |
switch_id | No (Required if `vpc_id` is configured) | The ID of the vSwitch in the VPC that the job accesses. The default value is empty. | STRING | Yes |
security_group_id | No (Required if `vpc_id` is configured) | The ID of the security group in the VPC that the job accesses. The default value is empty. | STRING | Yes |
job_file | No | The job parameter description file. Parameters in the `job_file` take precedence. The format is | STRING | No |
interactive | No | Specifies whether to start the job in interactive mode. | BOOL | Yes |
job_max_running_time_minutes | No | The maximum runtime of the job in minutes. The default value is 0, which means no maximum runtime is set. | INT64 | Yes |
success_policy | No | This parameter is supported only for TFJob. Valid values:
The default value is empty, which indicates AllWorkers. | STRING | Yes |
envs | No | The environment variables for the worker. The format is | StringToString | Yes |
tags | No | The tags of the job. The format is: | StringToString | Yes |
oversold_type | No | Specifies how the job uses idle resources. Valid values:
| STRING | Yes |
driver | No | The version of the GPU driver that the job uses. | STRING | Yes |
default_route | No | When a VPC is selected, this parameter specifies how to access the public network. Valid values:
| STRING | Yes |
priority | No | The priority of the job. The default value is 1. The value ranges from 1 to 9. Where:
| INT32 | Yes |
exit_code_on_stopped | No | When a job runs in interactive mode, this parameter specifies the exit code of the command line interface when the job is stopped. The default value is 0. | INT32 | Yes |
job_reserved_minutes | No | The retention duration of the job after it ends, in minutes. The default value is 0. | INT32 | Yes |
job_reserved_policy | No | The retention policy for the job. Valid values:
| STRING | Yes |
Submit TensorFlow training jobs (submit tfjob)
Feature
You can use this command to submit TensorFlow training jobs.
Format
You can submit TensorFlow jobs using command line parameters or a job parameter description file.
./dlc submit tfjob [flags]Parameters
When you submit a TensorFlow job using command line parameters, replace the parameters in the command with your actual values. When you submit a job using a job parameter description file, specify the parameters in the file in the
<parameterName>=<parameterValue>format. For information about common parameters, see the list at the beginning of this topic. The following table describes the parameters that are specific to TensorFlow jobs.Table 2. Parameters specific to TensorFlow jobs
Parameter
Required
Description
Type
Supported in job parameter description file
workspace_id
Yes
The ID of the workspace. The default value is empty. For more information about how to create a workspace, see Create and manage workspaces.
STRING
Yes
chief
No
Specifies whether to enable the TensorFlow Chief node. Valid values:
false: (Default) Disables the TensorFlow Chief node.
true: Enables the TensorFlow Chief node.
BOOL
Yes
chief_image
No
The runtime image for the TensorFlow Chief node. The default value is empty.
STRING
Yes
chief_spec
No
The server model used by the TensorFlow Chief node. The default value is empty.
STRING
Yes
master_image
No
The runtime image for the TensorFlow master node. The default value is empty.
STRING
Yes
master_spec
No
The server model used by the TensorFlow master node.
STRING
Yes
masters
No
The number of TensorFlow master nodes. The default value is 0.
INT
Yes
ps
No
The number of TensorFlow Parameter Server nodes. The default value is 0.
INT
Yes
ps_image
No
The runtime image for the TensorFlow Parameter Server node. The default value is empty.
STRING
Yes
ps_spec
No
The server model used by the TensorFlow Parameter Server node. The default value is empty.
STRING
Yes
worker_image
No
The runtime image for the TensorFlow Worker node. The default value is empty.
STRING
Yes
worker_spec
No
The server model used by the TensorFlow Worker node. The default value is empty.
STRING
Yes
workers
No
The number of TensorFlow Worker nodes. The default value is 0.
INT
Yes
evaluator_image
No
The runtime image for the TensorFlow Evaluator node. The default value is empty.
STRING
Yes
evaluator_spec
No
The server model used by the TensorFlow Evaluator node. The default value is empty.
STRING
Yes
evaluators
No
The number of TensorFlow Evaluator nodes. The default value is 0.
INT
Yes
graphlearn_image
No
The runtime image for the TensorFlow GraphLearn node. The default value is empty.
STRING
Yes
graphlearn_spec
No
The server model used by the TensorFlow GraphLearn node. The default value is empty.
STRING
Yes
graphlearns
No
The number of TensorFlow GraphLearn nodes. The default value is 0.
INT
Yes
Table 3. Parameters specific to submitting TensorFlow jobs to a dedicated resource group
Parameter
Required
Description
Type
Supported in job parameter description file
resource_id
No (Required if you want to submit a job to a dedicated resource group)
The ID of the dedicated resource quota. The default value is empty. For more information about how to create a dedicated resource quota, see General computing resource quotas.
STRING
Yes
priority
No
The job priority. The default value is 1.
INT
Yes
chief_cpu
No
The number of CPUs for the TensorFlow Chief node. The default value is empty.
STRING
Yes
chief_gpu
No
The number of GPUs for the TensorFlow Chief node. The default value is empty.
STRING
Yes
chief_gpu_type
No
The type of GPU for the TensorFlow Chief node. The default value is empty. Example: GU50.
STRING
Yes
chief_memory
No
The memory resources for the TensorFlow Chief node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
chief_shared_memory
No
The shared memory resources for the TensorFlow Chief node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
master_cpu
No
The number of CPUs for the TensorFlow master node. The default value is empty.
STRING
Yes
master_gpu
No
The number of GPUs for the TensorFlow master node. The default value is empty.
STRING
Yes
master_gpu_type
No
The type of GPU for the TensorFlow master node. The default value is empty. Example: GU50.
STRING
Yes
master_memory
No
The memory resources for the TensorFlow master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
master_shared_memory
No
The shared memory resources for the TensorFlow master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
*_cpu
No
The number of CPUs for the TensorFlow * node. The default value is empty. * can be replaced with `ps`, `worker`, `evaluator`, or `graphlearn`.
STRING
Yes
*_gpu
No
The number of GPUs for the TensorFlow * node. The default value is empty. * can be replaced with `ps`, `worker`, `evaluator`, or `graphlearn`.
STRING
Yes
*_gpu_type
No
The type of GPU for the TensorFlow * node. The default value is empty. Example: GU50. * can be replaced with `ps`, `worker`, `evaluator`, or `graphlearn`.
STRING
Yes
*_memory
No
The memory resources for the TensorFlow * node. The default value is empty. Examples: 500 Mi, 1 Gi. * can be replaced with `ps`, `worker`, `evaluator`, or `graphlearn`.
STRING
Yes
*_shared_memory
No
The shared memory resources for the TensorFlow * node. The default value is empty. Examples: 500 Mi, 1 Gi. * can be replaced with `ps`, `worker`, `evaluator`, or `graphlearn`.
STRING
Yes
Examples
The following example shows how to submit a distributed job with two Worker nodes and one PS node using command line parameters.
./dlc submit tfjob --name=test_2021 --ps=1 \ --ps_spec=ecs.g6.8xlarge \ --ps_image=registry-vpc.cn-beijing.aliyuncs.com/pai-dlc/tensorflow-training:1.12.2PAI-cpu-py27-ubuntu16.04 \ --workers=2 \ --worker_spec=ecs.g6.4xlarge \ --worker_image=registry-vpc.cn-beijing.aliyuncs.com/pai-dlc/tensorflow-training:1.12.2PAI-cpu-py27-ubuntu16.04 \ --command="python /root/data/dist_mnist/code/dist-main.py --max_steps=10000 --data_dir=/root/data/dist_mnist/data/" \ --workspace_id=***** \ --data_sources=data-2021xxxxxxxxxx-xxxxxxxxxxxxThe following result is returned.
+----------------------------------+--------------------------------------+ | JobId | RequestId | +----------------------------------+--------------------------------------+ | dlcmp6vwljkz**** | xxxxxxxx-79AF-4EFC-9CE9-xxxxxxxxxxxx | +----------------------------------+--------------------------------------+The following example shows how to submit a distributed job with two Worker nodes and one PS node using a job parameter description file.
./dlc submit tfjob --job_file=job_file.dist_mnist.1ps2wIn the command, job_file.dist_mnist.1ps2w is the job parameter description file. The parameters are specified in the
<parameterName>=<parameterValue>format. The following code provides an example of the content in job_file.dist_mnist.1ps2w.name=test_2021 workers=2 worker_spec=ecs.g6.4xlarge worker_image=registry-vpc.cn-beijing.aliyuncs.com/pai-dlc/tensorflow-training:1.12.2PAI-cpu-py27-ubuntu16.04 ps=1 ps_spec=ecs.g6.8xlarge ps_image=registry-vpc.cn-beijing.aliyuncs.com/pai-dlc/tensorflow-training:1.12.2PAI-cpu-py27-ubuntu16.04 command=python /root/data/dist_mnist/code/dist-main.py --max_steps=10000 --data_dir=/root/data/dist_mnist/data/ workspace_id=***** data_sources=data-2021xxxxxxxxxx-xxxxxxxxxxxx
Submit PyTorch training jobs (submit pytorchjob)
Feature
You can use this command to submit PyTorch training jobs.
Format
You can submit PyTorch jobs using command line parameters or a job parameter description file.
./dlc submit pytorchjob [flags]Parameters
When you submit a PyTorch job using command line parameters, replace the parameters in the command with your actual values. When you submit a job using a job parameter description file, specify the parameters in the file in the
<parameterName>=<parameterValue>format. For information about common parameters, see the list at the beginning of this topic. The following table describes the parameters that are specific to PyTorch jobs.Table 4. Parameters specific to PyTorch jobs
Parameter name
Required
Description
Type
Supported in job parameter description file
workspace_id
Yes
The ID of the workspace. The default value is empty. For more information about how to create a workspace, see Create and manage workspaces.
STRING
Yes
master_image
No
The runtime image for the PyTorch master node. The default value is empty.
STRING
Yes
master_spec
No
The server model used by the PyTorch master node. The default value is empty.
STRING
Yes
masters
No
The number of PyTorch master nodes. The default value is 0.
INT
Yes
worker_image
No
The runtime image for the PyTorch Worker node. The default value is empty.
STRING
Yes
worker_spec
No
The server model used by the PyTorch Worker node. The default value is empty.
STRING
Yes
workers
No
The number of PyTorch Worker nodes. The default value is 0.
INT
Yes
Table 5. Parameters specific to submitting PyTorch jobs to a dedicated resource group
Parameter
Required
Description
Type
Supported in job parameter description file
resource_id
No (Required if you want to submit a job to a dedicated resource group)
The ID of the dedicated resource quota. The default value is empty. For more information about how to create a dedicated resource quota, see General computing resource quotas.
STRING
Yes
priority
No
The job priority. The default value is 1.
INT
Yes
master_cpu
No
The number of CPUs for the PyTorch master node. The default value is empty.
STRING
Yes
master_gpu
No
The number of GPUs for the PyTorch master node. The default value is empty.
STRING
Yes
master_gpu_type
No
The type of GPU for the PyTorch master node. The default value is empty. Example: GU50.
STRING
Yes
master_memory
No
The memory resources for the PyTorch master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
master_shared_memory
No
The shared memory resources for the PyTorch master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
worker_cpu
No
The number of CPUs for the PyTorch Worker node. The default value is empty.
STRING
Yes
worker_gpu
No
The number of GPUs for the PyTorch Worker node. The default value is empty.
STRING
Yes
worker_gpu_type
No
The type of GPU for the PyTorch Worker node. The default value is empty. Example: GU50.
STRING
Yes
worker_memory
No
The memory resources for the PyTorch Worker node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
worker_shared_memory
No
The shared memory resources for the PyTorch Worker node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
Example
The following example shows how to submit a GPU model training job using command line parameters.
./dlc submit pytorchjob --name=test_pt_face \ --workers=1 \ --worker_spec=ecs.gn6e-c12g1.3xlarge \ --worker_image=registry-vpc.cn-beijing.aliyuncs.com/pai-dlc/pytorch-training:1.7.1-gpu-py37-cu110-ubuntu18.04 \ --command="apt-get update; apt-get -y --allow-downgrades install libpcre3=2:8.38-3.1 libpcre3-dev libgl1-mesa-glx libglib2.0-dev; cd /root/data/face; python train.py --num_workers 0 --save_folder outputs" \ --data_sources=data-20210410224621-xxxxxxxxxxxx \ --workspace_id=*****The following result is returned.
+----------------------------------+--------------------------------------+ | JobId | RequestId | +----------------------------------+--------------------------------------+ | dlcu704xxuxk**** | xxxxxxxx-79AF-4EFC-9CE9-xxxxxxxxxxxx | +----------------------------------+--------------------------------------+
Submit XGBoost training jobs (submit xgboostjob)
Feature
You can use this command to submit XGBoost training jobs.
Format
You can submit XGBoost jobs using command line parameters or a job parameter description file.
./dlc submit xgboostjob [flags]Parameters
When you submit an XGBoost job using command line parameters, replace the parameters in the command with your actual values. When you submit a job using a job parameter description file, specify the parameters in the file in the
<parameterName>=<parameterValue>format. For information about common parameters, see the list at the beginning of this topic. The following table describes the parameters that are specific to XGBoost jobs.Table 6. Parameters specific to XGBoost jobs
Parameter name
Required
Description
Type
Supported in job parameter description file
workspace_id
Yes
The ID of the workspace. The default value is empty. For more information about how to create a workspace, see Create and manage workspaces.
STRING
Yes
master_image
No
The runtime image for the XGBoost master node. The default value is empty.
STRING
Yes
master_spec
No
The server model used by the XGBoost master node. The default value is empty.
STRING
Yes
masters
No
The number of XGBoost master nodes. The default value is 0.
INT
Yes
worker_image
No
The runtime image for the XGBoost Worker node. The default value is empty.
STRING
Yes
worker_spec
No
The server model used by the XGBoost Worker node. The default value is empty.
STRING
Yes
workers
No
The number of XGBoost Worker nodes. The default value is 0.
INT
Yes
Table 7. Parameters specific to submitting XGBoost jobs to a dedicated resource group
Parameter
Required
Description
Type
Supported in job parameter description file
resource_id
No (Required if you want to submit a job to a dedicated resource group)
The ID of the dedicated resource quota. The default value is empty. For more information about how to create a dedicated resource quota, see General computing resource quotas.
STRING
Yes
priority
No
The job priority. The default value is 1.
INT
Yes
master_cpu
No
The number of CPUs for the XGBoost master node. The default value is empty.
STRING
Yes
master_gpu
No
The number of GPUs for the XGBoost master node. The default value is empty.
STRING
Yes
master_gpu_type
No
The type of GPU for the XGBoost master node. The default value is empty. Example: GU50.
STRING
Yes
master_memory
No
The memory resources for the XGBoost master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
master_shared_memory
No
The shared memory resources for the XGBoost master node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
worker_cpu
No
The number of CPUs for the XGBoost Worker node. The default value is empty.
STRING
Yes
worker_gpu
No
The number of GPUs for the XGBoost Worker node. The default value is empty.
STRING
Yes
worker_gpu_type
No
The type of GPU for the XGBoost Worker node. The default value is empty. Example: GU50.
STRING
Yes
worker_memory
No
The memory resources for the XGBoost Worker node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
worker_shared_memory
No
The shared memory resources for the XGBoost Worker node. The default value is empty. Examples: 500 Mi, 1 Gi.
STRING
Yes
Example
The following example shows how to submit an XGBoost job using command line parameters.
./dlc submit xgboostjob --name=test_xgboost \ --workers=1 \ --worker_spec=ecs.gn6e-c12g1.3xlarge \ --worker_image=xgboost-training:1.6.0-cpu-py36-ubuntu18.04 \ --command="python /root/code/horovod/xgboost/main.py --job_type=Train --xgboost_parameter=objective:multi:softprob,num_class:3 --n_estimators=50 --model_path=autoAI/xgb-opt/2" \ --workspace_id=*****The following result is returned.
+----------------------------------+--------------------------------------+ | JobId | RequestId | +----------------------------------+--------------------------------------+ | dlc1nvu3gli0**** | xxxxxxxx-79AF-4EFC-9CE9-xxxxxxxxxxxx | +----------------------------------+--------------------------------------+
Advanced parameters for submitting jobs
Use specified node scheduling when submitting a job
When you submit a job using a resource quota for Lingjun AI Computing or general computing, you can configure parameters in the DLC command line interface to schedule specific nodes.
This feature is available only to users on the whitelist. To use this feature, contact your account manager to add your account to the whitelist.
Parameters
Parameter
Description
Example value
--allow_nodes="${allow_nodes}"
A list of node names. Separate multiple node names with commas (,). We recommend that you do not add spaces between the names.
lingjuc47iextvg9-***,lingjuc47iextvg9-***
--deny_nodes="${deny_nodes}"
A list of node names to exclude. Separate multiple node names with commas (,). We recommend that you do not add spaces between the names.
lingjuc47iextvg9-***,lingjuc47iextvg9-***
Examples
Command line parameters
The following examples show how to submit a job using command line parameters:
Do not specify a scheduling node
./dlc submit pytorchjob --name=assign_node_test_no_node \--workers=1 \ --worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 \ --command="sleep 1000" \ --workspace_id='****' \ --resource_id='quotau2h98mt****' \ --worker_cpu="1" \ --worker_memory='2Gi'Specify scheduling nodes
./dlc submit pytorchjob --name=assign_node_test_2_allow_nodes \--workers=1 \ --worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 \ --command="sleep 1000" \ --workspace_id='****' \ --resource_id='quotau2h98mt****' \ --worker_cpu="1" \ --worker_memory='2Gi' \ --allow_nodes="lingjuc47iextvg9-****,lingjuc47iextvg9-****"Exclude specified nodes
./dlc submit pytorchjob --name=assign_node_test_two_deny_nodes \--workers=1 \ --worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 \ --command="sleep 1000" \ --workspace_id='****' \ --resource_id='quotau2h98mt****' \ --worker_cpu="1" \ --worker_memory='2Gi' \ --deny_nodes="lingjuc47iextvg9-****,lingjuc47iextvg9-****"Specify and exclude nodes
./dlc submit pytorchjob --name=assign_node_test_two_allow_two_deny \--workers=1 \ --worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 \ --command="sleep 1000" \ --workspace_id='****' \ --resource_id='quotau2h98mt****' \ --worker_cpu="1" \ --worker_memory='2Gi' \ --allow_nodes="lingjuc47iextvg9-****,lingjuc47iextvg9-****" \ --deny_nodes="lingjuc47iextvg9-****,lingjuc47iextvg9-****"
Read from a file
To submit a job by reading from a file, use the following command:
./dlc submit pytorchjob -f job_fileIn the command, job_file is the job parameter description file. The following code provides an example of the file content:
Do not specify a scheduling node
name=assign_node_test_no_node workers=1 worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 command=sleep 1000 workspace_id=**** resource_id=quotau2h98mt**** worker_cpu=1 worker_memory=2GiSpecify scheduling nodes
name=assign_node_test_2_allow_nodes workers=1 worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 command=sleep 1000 workspace_id=**** resource_id=quotau2h98mt**** worker_cpu=1 worker_memory=2Gi allow_nodes=lingjuc47iextvg9-****,lingjuc47iextvg9-****Exclude specified nodes
name=assign_node_test_two_allow_two_deny workers=1 worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 command=sleep 1000 workspace_id=**** resource_id=quotau2h98mt**** worker_cpu=1 worker_memory=2Gi deny_nodes=lingjuc47iextvg9-****,lingjuc47iextvg9-****Specify and exclude nodes
name=assign_node_test_two_allow_two_deny workers=1 worker_image=dsw-registry-vpc.****.cr.aliyuncs.com/pai/easyanimate:1.1.5-pytorch2.2.0-gpu-py310-cu118-ubuntu22.04 command=sleep 1000 workspace_id=**** resource_id=quotau2h98mt**** worker_cpu=1 worker_memory=2Gi allow_nodes=lingjuc47iextvg9-****,lingjuc47iextvg9-**** deny_nodes=lingjuc47iextvg9-****,lingjuc47iextvg9-****
Disable pay-as-you-go inventory check when submitting a job
You can configure the disable_ecs_stock_check parameter in the DLC command line interface to disable the pay-as-you-go inventory check feature.
Parameter
Parameter
Description
Example
disable_ecs_stock_check
Disables the pay-as-you-go inventory check. Valid values:
false (default): Enables the pay-as-you-go inventory check.
true: Disables the pay-as-you-go inventory check.
true or false
Examples
Command line parameters
The following examples show how to submit a job using command line parameters:
Enable the pay-as-you-go inventory check
./dlc submit pytorchjob \ --name=test_skip_checking3 \ --command='sleep 1000' \ --workspace_id=**** \ --priority=1 \ --workers=1 \ --worker_image=registry.cn-hangzhou.aliyuncs.com/pai-dlc/tensorflow-training:1.12PAI-gpu-py36-cu101-ubuntu18.04 \ --worker_spec=ecs.g6.xlargeDisable the pay-as-you-go inventory check
./dlc submit pytorchjob \ --name=test_skip_checking3 \ --command='sleep 1000' \ --workspace_id=**** \ --priority=1 \ --workers=1 \ --worker_image=registry.cn-hangzhou.aliyuncs.com/pai-dlc/tensorflow-training:1.12PAI-gpu-py36-cu101-ubuntu18.04 \ --worker_spec=ecs.g6.xlarge \ --disable_ecs_stock_check=true
Read from a file
To submit a job by reading from a file, use the following command:
./dlc submit pytorchjob -f job_fileIn the command, job_file is the job parameter description file. The following code provides an example of the file content:
Enable the pay-as-you-go inventory check
name=test_skip_checking3 workers=1 worker_image=registry.cn-hangzhou.aliyuncs.com/pai-dlc/tensorflow-training:1.12PAI-gpu-py36-cu101-ubuntu18.04 command=sleep 1000 workspace_id=**** worker_spec=ecs.g6.xlargeDisable the pay-as-you-go inventory check
name=test_skip_checking3 workers=1 worker_image=registry.cn-hangzhou.aliyuncs.com/pai-dlc/tensorflow-training:1.12PAI-gpu-py36-cu101-ubuntu18.04 command=sleep 1000 workspace_id=**** worker_spec=ecs.g6.xlarge disable_ecs_stock_check=true
References
After you submit a job, you can manage the job using the DLC client. For more information, see Stop commands and Query commands.
You can also manage submitted jobs in the console. For more information, see Manage training jobs.