CloudAI Benchmark Framework v1.5.0

Bash Command

This workload (test_template_name is BashCmd) allows users to execute arbitrary bash commands within the CloudAI framework. This is useful for simple scripts, custom testing commands, or integrating external tools.

cmd specified in the cmd_args section will be added as-is into generated sbatch script.

Test TOML example:

Copy
Copied!
            

name = "my_bash_test" description = "Example bash command test" test_template_name = "BashCmd" [cmd_args] cmd = "echo 'Hello from CloudAI!'"

Test Scenario example:

Copy
Copied!
            

name = "bash-test" [[Tests]] id = "bash.1" num_nodes = 1 time_limit = "00:05:00" test_name = "my_bash_test"

Test-in-Scenario example:

Copy
Copied!
            

name = "bash-test" [[Tests]] id = "bash.1" num_nodes = 1 time_limit = "00:05:00" name = "my_bash_test" description = "Example bash command test" test_template_name = "BashCmd" [Tests.cmd_args] cmd = "echo 'Hello from CloudAI!'"

Command Arguments

class cloudai.workloads.bash_cmd.bash_cmd.BashCmdArgs(*, cmd: str, **extra_data: Any)[source]

Bases: CmdArgs

Arguments for a Bash command.

cmd: str

Test Definition

class cloudai.workloads.bash_cmd.bash_cmd.BashCmdTestDefinition(*, name: str, description: str, test_template_name: str, cmd_args: BashCmdArgs, extra_env_vars: dict[str, str | List[str]] = {}, extra_cmd_args: dict[str, str] = {}, extra_container_mounts: list[str] = [], git_repos: list[GitRepo] = [], nsys: NsysConfiguration | None = None, predictor: PredictorConfig | None = None, agent: str = 'grid_search', agent_steps: int = 1, agent_metrics: list[str] = ['default'], agent_reward_function: str = 'inverse')[source]

Bases: TestDefinition

Test definition for a Bash command.

cmd_args: BashCmdArgs
property installables: list[Installable]
Previous AIConfigurator
Next Chakra Replay
© Copyright 2026, NVIDIA CORPORATION & AFFILIATES. Last updated on Mar 3, 2026