diff --git a/docs/Researcher/cli-reference/new-cli/runai_config_set.md b/docs/Researcher/cli-reference/new-cli/runai_config_set.md index 4aa091a223..f2da1869ad 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_config_set.md +++ b/docs/Researcher/cli-reference/new-cli/runai_config_set.md @@ -6,14 +6,22 @@ Set configuration values runai config set [flags] ``` +### Examples + +``` +runai config set --status-timeout-duration 5s +runai config set --status-timeout-duration 300ms +``` + ### Options ``` - --auth-url string set the authorization URL; most likely the same as the control plane URL - --cp-url string set the control plane URL - -h, --help help for set - --interactive enable set interactive mode (enabled|disabled) - --output string set the default output type + --auth-url string set the authorization URL; most likely the same as the control plane URL + --cp-url string set the control plane URL + -h, --help help for set + --interactive enable set interactive mode (enabled|disabled) + --output string set the default output type + --status-timeout-duration string set cluster status call timeout duration value, the default is 3 second ("3s") ``` ### Options inherited from parent commands diff --git a/docs/Researcher/cli-reference/new-cli/runai_mpi.md b/docs/Researcher/cli-reference/new-cli/runai_mpi.md index 55df6b57c0..f944b0170e 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_mpi.md +++ b/docs/Researcher/cli-reference/new-cli/runai_mpi.md @@ -22,6 +22,7 @@ alias for mpi management * [runai](runai.md) - Run:ai Command-line Interface * [runai mpi attach](runai_mpi_attach.md) - attach to a running container in a mpi training job +* [runai mpi bash](runai_mpi_bash.md) - open a bash shell in a training mpi job * [runai mpi delete](runai_mpi_delete.md) - delete mpi training workload * [runai mpi describe](runai_mpi_describe.md) - describe mpi training * [runai mpi exec](runai_mpi_exec.md) - execute a command in a training mpi job diff --git a/docs/Researcher/cli-reference/new-cli/runai_mpi_bash.md b/docs/Researcher/cli-reference/new-cli/runai_mpi_bash.md new file mode 100644 index 0000000000..419a7a5808 --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_mpi_bash.md @@ -0,0 +1,45 @@ +## runai mpi bash + +open a bash shell in a training mpi job + +``` +runai mpi bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training mpi's main worker +runai training mpi bash mpi-01 + +# Open a bash shell in a specific training mpi worker +runai training mpi bash mpi-01 --pod mpi-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai mpi](runai_mpi.md) - alias for mpi management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_mpi_submit.md b/docs/Researcher/cli-reference/new-cli/runai_mpi_submit.md index b6c64f44aa..b8c4a7cbee 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_mpi_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_mpi_submit.md @@ -52,7 +52,7 @@ runai training mpi submit -p -i gcr.io/run-ai-demo/quickst --master-environment stringArray Set master environment variables in the container --master-extended-resource stringArray Request access to an extended resource. Use the format: resource_name=quantity --master-no-pvcs Do not mount any persistent volumes in the master pod - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_pytorch.md b/docs/Researcher/cli-reference/new-cli/runai_pytorch.md index 1b9b0a7595..b979437f68 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_pytorch.md +++ b/docs/Researcher/cli-reference/new-cli/runai_pytorch.md @@ -22,6 +22,7 @@ alias for pytorch management * [runai](runai.md) - Run:ai Command-line Interface * [runai pytorch attach](runai_pytorch_attach.md) - attach to a running container in a pytorch training job +* [runai pytorch bash](runai_pytorch_bash.md) - open a bash shell in a training pytorch job * [runai pytorch delete](runai_pytorch_delete.md) - delete pytorch training workload * [runai pytorch describe](runai_pytorch_describe.md) - describe pytorch training * [runai pytorch exec](runai_pytorch_exec.md) - execute a command in a training pytorch job diff --git a/docs/Researcher/cli-reference/new-cli/runai_pytorch_bash.md b/docs/Researcher/cli-reference/new-cli/runai_pytorch_bash.md new file mode 100644 index 0000000000..21bf9bb0fe --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_pytorch_bash.md @@ -0,0 +1,45 @@ +## runai pytorch bash + +open a bash shell in a training pytorch job + +``` +runai pytorch bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training pytorch's main worker +runai training pytorch bash pytorch-01 + +# Open a bash shell in a specific training pytorch worker +runai training pytorch bash pytorch-01 --pod pytorch-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai pytorch](runai_pytorch.md) - alias for pytorch management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_pytorch_submit.md b/docs/Researcher/cli-reference/new-cli/runai_pytorch_submit.md index 3acef98784..c371d048b6 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_pytorch_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_pytorch_submit.md @@ -56,7 +56,7 @@ runai training pytorch submit -p -i gcr.io/run-ai-demo/qui --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod --max-replicas int32 Maximum number of replicas for an elastic PyTorch job - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --min-replicas int32 Minimum number of replicas for an elastic PyTorch job --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral diff --git a/docs/Researcher/cli-reference/new-cli/runai_submit.md b/docs/Researcher/cli-reference/new-cli/runai_submit.md index 17681d1669..ef444ed901 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_submit.md @@ -39,7 +39,7 @@ runai submit [flags] --large-shm Request large /dev/shm device to mount --memory string CPU memory to allocate for the job (e.g. 1G, 500M) --memory-limit string CPU memory limit to allocate for the job (e.g. 1G, 500M) - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --node-pools stringArray List of node pools to use for scheduling the job, ordered by priority --node-type string Enforce node type affinity by setting a node-type label diff --git a/docs/Researcher/cli-reference/new-cli/runai_tensorflow.md b/docs/Researcher/cli-reference/new-cli/runai_tensorflow.md index 4bdc3ed4c0..240761b323 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_tensorflow.md +++ b/docs/Researcher/cli-reference/new-cli/runai_tensorflow.md @@ -22,6 +22,7 @@ alias for tensorflow management * [runai](runai.md) - Run:ai Command-line Interface * [runai tensorflow attach](runai_tensorflow_attach.md) - attach to a running container in a tf training job +* [runai tensorflow bash](runai_tensorflow_bash.md) - open a bash shell in a training tf job * [runai tensorflow delete](runai_tensorflow_delete.md) - delete tf training workload * [runai tensorflow describe](runai_tensorflow_describe.md) - describe tf training * [runai tensorflow exec](runai_tensorflow_exec.md) - execute a command in a training tf job diff --git a/docs/Researcher/cli-reference/new-cli/runai_tensorflow_bash.md b/docs/Researcher/cli-reference/new-cli/runai_tensorflow_bash.md new file mode 100644 index 0000000000..8624ff784f --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_tensorflow_bash.md @@ -0,0 +1,45 @@ +## runai tensorflow bash + +open a bash shell in a training tf job + +``` +runai tensorflow bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training tf's main worker +runai training tf bash tf-01 + +# Open a bash shell in a specific training tf worker +runai training tf bash tf-01 --pod tf-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai tensorflow](runai_tensorflow.md) - alias for tensorflow management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_tensorflow_submit.md b/docs/Researcher/cli-reference/new-cli/runai_tensorflow_submit.md index 4d3b97464c..5a634cd2ad 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_tensorflow_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_tensorflow_submit.md @@ -56,7 +56,7 @@ runai training tf submit -p -i gcr.io/run-ai-demo/quicksta --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod --max-replicas int32 Maximum number of replicas for an elastic PyTorch job - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --min-replicas int32 Minimum number of replicas for an elastic PyTorch job --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral diff --git a/docs/Researcher/cli-reference/new-cli/runai_training.md b/docs/Researcher/cli-reference/new-cli/runai_training.md index 0f0803b872..40faf72a1e 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training.md @@ -22,10 +22,11 @@ training management * [runai](runai.md) - Run:ai Command-line Interface * [runai training attach](runai_training_attach.md) - attach to a running container in a standard training job +* [runai training bash](runai_training_bash.md) - open a bash shell in a training standard job * [runai training delete](runai_training_delete.md) - delete standard training workload * [runai training describe](runai_training_describe.md) - describe standard training * [runai training exec](runai_training_exec.md) - execute a command in a training standard job -* [runai training list](runai_training_list.md) - list training +* [runai training list](runai_training_list.md) - list all training frameworks * [runai training logs](runai_training_logs.md) - view logs of a standard training job * [runai training mpi](runai_training_mpi.md) - mpi management * [runai training port-forward](runai_training_port-forward.md) - forward one or more local ports to a standard training job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_bash.md new file mode 100644 index 0000000000..af7a318b21 --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_bash.md @@ -0,0 +1,45 @@ +## runai training bash + +open a bash shell in a training standard job + +``` +runai training bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training standard's main worker +runai training standard bash standard-01 + +# Open a bash shell in a specific training standard worker +runai training standard bash standard-01 --pod standard-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training](runai_training.md) - training management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_list.md b/docs/Researcher/cli-reference/new-cli/runai_training_list.md index e5fc8a2761..ca540d7155 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_list.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_list.md @@ -1,6 +1,6 @@ ## runai training list -list training +list all training frameworks ``` runai training list [flags] diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_mpi.md b/docs/Researcher/cli-reference/new-cli/runai_training_mpi.md index cd56cbca1b..592c3f2740 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_mpi.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_mpi.md @@ -22,6 +22,7 @@ mpi management * [runai training](runai_training.md) - training management * [runai training mpi attach](runai_training_mpi_attach.md) - attach to a running container in a mpi training job +* [runai training mpi bash](runai_training_mpi_bash.md) - open a bash shell in a training mpi job * [runai training mpi delete](runai_training_mpi_delete.md) - delete mpi training workload * [runai training mpi describe](runai_training_mpi_describe.md) - describe mpi training * [runai training mpi exec](runai_training_mpi_exec.md) - execute a command in a training mpi job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_mpi_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_mpi_bash.md new file mode 100644 index 0000000000..bb588ed776 --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_mpi_bash.md @@ -0,0 +1,45 @@ +## runai training mpi bash + +open a bash shell in a training mpi job + +``` +runai training mpi bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training mpi's main worker +runai training mpi bash mpi-01 + +# Open a bash shell in a specific training mpi worker +runai training mpi bash mpi-01 --pod mpi-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training mpi](runai_training_mpi.md) - mpi management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_mpi_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_mpi_submit.md index 6ada4fe58d..f3eb42bffb 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_mpi_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_mpi_submit.md @@ -52,7 +52,7 @@ runai training mpi submit -p -i gcr.io/run-ai-demo/quickst --master-environment stringArray Set master environment variables in the container --master-extended-resource stringArray Request access to an extended resource. Use the format: resource_name=quantity --master-no-pvcs Do not mount any persistent volumes in the master pod - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_pytorch.md b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch.md index b6ba8c23fc..2ac9e4bbc8 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_pytorch.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch.md @@ -22,6 +22,7 @@ pytorch management * [runai training](runai_training.md) - training management * [runai training pytorch attach](runai_training_pytorch_attach.md) - attach to a running container in a pytorch training job +* [runai training pytorch bash](runai_training_pytorch_bash.md) - open a bash shell in a training pytorch job * [runai training pytorch delete](runai_training_pytorch_delete.md) - delete pytorch training workload * [runai training pytorch describe](runai_training_pytorch_describe.md) - describe pytorch training * [runai training pytorch exec](runai_training_pytorch_exec.md) - execute a command in a training pytorch job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_bash.md new file mode 100644 index 0000000000..fcfb3cabdc --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_bash.md @@ -0,0 +1,45 @@ +## runai training pytorch bash + +open a bash shell in a training pytorch job + +``` +runai training pytorch bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training pytorch's main worker +runai training pytorch bash pytorch-01 + +# Open a bash shell in a specific training pytorch worker +runai training pytorch bash pytorch-01 --pod pytorch-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training pytorch](runai_training_pytorch.md) - pytorch management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_submit.md index d8959525ad..fee383d370 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_pytorch_submit.md @@ -56,7 +56,7 @@ runai training pytorch submit -p -i gcr.io/run-ai-demo/qui --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod --max-replicas int32 Maximum number of replicas for an elastic PyTorch job - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --min-replicas int32 Minimum number of replicas for an elastic PyTorch job --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_standard.md b/docs/Researcher/cli-reference/new-cli/runai_training_standard.md index 32b417c5fd..7507253f06 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_standard.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_standard.md @@ -22,6 +22,7 @@ standard training management * [runai training](runai_training.md) - training management * [runai training standard attach](runai_training_standard_attach.md) - attach to a running container in a standard training job +* [runai training standard bash](runai_training_standard_bash.md) - open a bash shell in a training standard job * [runai training standard delete](runai_training_standard_delete.md) - delete standard training workload * [runai training standard describe](runai_training_standard_describe.md) - describe standard training * [runai training standard exec](runai_training_standard_exec.md) - execute a command in a training standard job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_standard_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_standard_bash.md new file mode 100644 index 0000000000..1ee9f63b10 --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_standard_bash.md @@ -0,0 +1,45 @@ +## runai training standard bash + +open a bash shell in a training standard job + +``` +runai training standard bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training standard's main worker +runai training standard bash standard-01 + +# Open a bash shell in a specific training standard worker +runai training standard bash standard-01 --pod standard-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training standard](runai_training_standard.md) - standard training management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_standard_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_standard_submit.md index 4fa1c921f9..a4e60f59b3 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_standard_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_standard_submit.md @@ -49,7 +49,7 @@ runai training standard submit -p -i gcr.io/run-ai-demo/qu --image-pull-policy string Set image pull policy. One of: Always, IfNotPresent, Never. Defaults to Always (default "Always") --label stringArray Set of labels to populate into the container running the workspace --large-shm Request large /dev/shm device to mount - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_submit.md index 2cb90400ca..889744453c 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_submit.md @@ -49,7 +49,7 @@ runai training standard submit -p -i gcr.io/run-ai-demo/qu --image-pull-policy string Set image pull policy. One of: Always, IfNotPresent, Never. Defaults to Always (default "Always") --label stringArray Set of labels to populate into the container running the workspace --large-shm Request large /dev/shm device to mount - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow.md b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow.md index 543b0a0c61..eca700c640 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow.md @@ -22,6 +22,7 @@ tensorflow management * [runai training](runai_training.md) - training management * [runai training tensorflow attach](runai_training_tensorflow_attach.md) - attach to a running container in a tf training job +* [runai training tensorflow bash](runai_training_tensorflow_bash.md) - open a bash shell in a training tf job * [runai training tensorflow delete](runai_training_tensorflow_delete.md) - delete tf training workload * [runai training tensorflow describe](runai_training_tensorflow_describe.md) - describe tf training * [runai training tensorflow exec](runai_training_tensorflow_exec.md) - execute a command in a training tf job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_bash.md new file mode 100644 index 0000000000..e6441528ab --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_bash.md @@ -0,0 +1,45 @@ +## runai training tensorflow bash + +open a bash shell in a training tf job + +``` +runai training tensorflow bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training tf's main worker +runai training tf bash tf-01 + +# Open a bash shell in a specific training tf worker +runai training tf bash tf-01 --pod tf-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training tensorflow](runai_training_tensorflow.md) - tensorflow management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_submit.md index 1e7da053c1..ff93644550 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_tensorflow_submit.md @@ -56,7 +56,7 @@ runai training tf submit -p -i gcr.io/run-ai-demo/quicksta --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod --max-replicas int32 Maximum number of replicas for an elastic PyTorch job - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --min-replicas int32 Minimum number of replicas for an elastic PyTorch job --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_xgboost.md b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost.md index d915dbffc0..ac13bb7c43 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_xgboost.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost.md @@ -22,6 +22,7 @@ xgboost management * [runai training](runai_training.md) - training management * [runai training xgboost attach](runai_training_xgboost_attach.md) - attach to a running container in a xgboost training job +* [runai training xgboost bash](runai_training_xgboost_bash.md) - open a bash shell in a training xgboost job * [runai training xgboost delete](runai_training_xgboost_delete.md) - delete xgboost training workload * [runai training xgboost describe](runai_training_xgboost_describe.md) - describe xgboost training * [runai training xgboost exec](runai_training_xgboost_exec.md) - execute a command in a training xgboost job diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_bash.md b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_bash.md new file mode 100644 index 0000000000..f700e7061a --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_bash.md @@ -0,0 +1,45 @@ +## runai training xgboost bash + +open a bash shell in a training xgboost job + +``` +runai training xgboost bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training xgboost's main worker +runai training xgboost bash xgboost-01 + +# Open a bash shell in a specific training xgboost worker +runai training xgboost bash xgboost-01 --pod xgboost-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai training xgboost](runai_training_xgboost.md) - xgboost management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_submit.md b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_submit.md index 221f3a5dd2..73bcef087c 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_training_xgboost_submit.md @@ -55,7 +55,7 @@ runai training xgboost submit -p -i gcr.io/run-ai-demo/qui --master-gpu-portion-limit float GPU portion limit, must be no less than the gpu-memory-request (between 0 and 1, e.g. 0.5, 0.2) --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_workload_describe.md b/docs/Researcher/cli-reference/new-cli/runai_workload_describe.md index 1b96faeb85..adba289825 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_workload_describe.md +++ b/docs/Researcher/cli-reference/new-cli/runai_workload_describe.md @@ -13,6 +13,7 @@ runai workload describe WORKLOAD_NAME [flags] --containers Include container information in pods --event-limit int32 Limit the number of events displayed (-1 for no limit) (default 50) --events Show events information (default true) + --framework string filter by workload framework --general Show general information (default true) -h, --help help for describe --networks Show networks information (default true) diff --git a/docs/Researcher/cli-reference/new-cli/runai_workspace.md b/docs/Researcher/cli-reference/new-cli/runai_workspace.md index 840ac2298b..897fd32a6a 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_workspace.md +++ b/docs/Researcher/cli-reference/new-cli/runai_workspace.md @@ -23,6 +23,7 @@ workspace management * [runai](runai.md) - Run:ai Command-line Interface * [runai workspace attach](runai_workspace_attach.md) - Attach to a process that is already running inside an existing container. +* [runai workspace bash](runai_workspace_bash.md) - open a bash shell in a job * [runai workspace delete](runai_workspace_delete.md) - delete workspace * [runai workspace describe](runai_workspace_describe.md) - Describe a training workload * [runai workspace exec](runai_workspace_exec.md) - exec management diff --git a/docs/Researcher/cli-reference/new-cli/runai_workspace_bash.md b/docs/Researcher/cli-reference/new-cli/runai_workspace_bash.md new file mode 100644 index 0000000000..84259e4f3e --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_workspace_bash.md @@ -0,0 +1,45 @@ +## runai workspace bash + +open a bash shell in a job + +``` +runai workspace bash WORKSPACE_NAME [flags] +``` + +### Examples + +``` +# Open a bash shell in the main worker +runai workspace bash workspace-01 + +# Open a bash shell in a specific worker +runai workspace bash workspace-01 --pod workspace-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai workspace](runai_workspace.md) - workspace management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_workspace_submit.md b/docs/Researcher/cli-reference/new-cli/runai_workspace_submit.md index 5a223426fa..ce76a2340d 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_workspace_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_workspace_submit.md @@ -47,7 +47,7 @@ runai workspace submit -p= -i=gcr.io/run-ai-demo/ --image-pull-policy string Set image pull policy. One of: Always, IfNotPresent, Never. Defaults to Always (default "Always") --label stringArray Set of labels to populate into the container running the workspace --large-shm Request large /dev/shm device to mount - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite diff --git a/docs/Researcher/cli-reference/new-cli/runai_xgboost.md b/docs/Researcher/cli-reference/new-cli/runai_xgboost.md index 1b9b27ff2c..fd37ffeae7 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_xgboost.md +++ b/docs/Researcher/cli-reference/new-cli/runai_xgboost.md @@ -22,6 +22,7 @@ alias for xgboost management * [runai](runai.md) - Run:ai Command-line Interface * [runai xgboost attach](runai_xgboost_attach.md) - attach to a running container in a xgboost training job +* [runai xgboost bash](runai_xgboost_bash.md) - open a bash shell in a training xgboost job * [runai xgboost delete](runai_xgboost_delete.md) - delete xgboost training workload * [runai xgboost describe](runai_xgboost_describe.md) - describe xgboost training * [runai xgboost exec](runai_xgboost_exec.md) - execute a command in a training xgboost job diff --git a/docs/Researcher/cli-reference/new-cli/runai_xgboost_bash.md b/docs/Researcher/cli-reference/new-cli/runai_xgboost_bash.md new file mode 100644 index 0000000000..acb8394296 --- /dev/null +++ b/docs/Researcher/cli-reference/new-cli/runai_xgboost_bash.md @@ -0,0 +1,45 @@ +## runai xgboost bash + +open a bash shell in a training xgboost job + +``` +runai xgboost bash [WORKLOAD_NAME] [flags] +``` + +### Examples + +``` +# Open a bash shell in the training xgboost's main worker +runai training xgboost bash xgboost-01 + +# Open a bash shell in a specific training xgboost worker +runai training xgboost bash xgboost-01 --pod xgboost-01-worker-1 +``` + +### Options + +``` + -c, --container string Container name for log extraction + -h, --help help for bash + --pod string Workload pod ID for log extraction, default: master (0-0) + --pod-running-timeout duration Pod check for running state timeout. + -p, --project string Specify the project to which the command applies. By default, commands apply to the default project. To change the default project use ‘runai config project ’ + -i, --stdin Pass stdin to the container + -t, --tty Stdin is a TTY + --wait-timeout duration Timeout for waiting for workload to be ready for log streaming +``` + +### Options inherited from parent commands + +``` + --config-file string config file name; can be set by environment variable RUNAI_CLI_CONFIG_FILE (default "config.json") + --config-path string config path; can be set by environment variable RUNAI_CLI_CONFIG_PATH (default "~/.runai/") + -d, --debug enable debug mode + -q, --quiet enable quiet mode, suppress all output except error messages + --verbose enable verbose mode +``` + +### SEE ALSO + +* [runai xgboost](runai_xgboost.md) - alias for xgboost management + diff --git a/docs/Researcher/cli-reference/new-cli/runai_xgboost_submit.md b/docs/Researcher/cli-reference/new-cli/runai_xgboost_submit.md index fedb177458..950a7df652 100644 --- a/docs/Researcher/cli-reference/new-cli/runai_xgboost_submit.md +++ b/docs/Researcher/cli-reference/new-cli/runai_xgboost_submit.md @@ -55,7 +55,7 @@ runai training xgboost submit -p -i gcr.io/run-ai-demo/qui --master-gpu-portion-limit float GPU portion limit, must be no less than the gpu-memory-request (between 0 and 1, e.g. 0.5, 0.2) --master-gpu-portion-request float GPU portion request (between 0 and 1, e.g. 0.5, 0.2) --master-no-pvcs Do not mount any persistent volumes in the master pod - --mig-profile string MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) + --mig-profile string [Deprecated] MIG profile to allocate for the job (1g.5gb, 2g.10gb, 3g.20gb, 4g.20gb, 7g.40gb) --name-prefix string Set defined prefix for the workload name and add index as a suffix --new-pvc stringArray Mount a persistent volume, create it if it does not exist. Use the format: claimname=CLAIM_NAME,storageclass=STORAGE_CLASS,size=SIZE,path=PATH,accessmode-rwo,accessmode-rom,accessmode-rwm,ro,ephemeral --nfs stringArray NFS storage details. Use the format: path=PATH,server=SERVER,mountpath=MOUNT_PATH,readwrite