> ## Documentation Index
> Fetch the complete documentation index at: https://tensorfuse.io/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# tensorkube job

## `tensorkube job`

Manage, deploy and queue jobs on Tensorfuse.

**Usage:**

```bash theme={null}
tensorkube job COMMAND [ARGS]...
```

**Options:**

* `--help BOOL` \[default: false]: Show this message and exit.

**Commands:**

* `deploy`: Deploy and orchestrate a job on Tenosrfuse.

* `delete`: Delete a job from Tensorfuse.

* `queue`: Queue an instance of a deployed job on Tensorfuse.

* `get`: Get the status of a job instance.

* `list`: List all jobs or a specific job on Tensorfuse.

* `logs`: Stream the logs for a specific job instance.

* `oneshot`: IN ALPHA; DO NOT USE. Manage and deploy one-shot jobs on Tensorfuse.

## `tensorkube job deploy`

Deploy and orchestrate a job on Tenosrfuse.

This command orchestrates the infrastructure needed for jobs such as queues, buckets and key-value stores.
It also builds the container image for the job and pushes it to the container registry.

**Usage:**

```bash theme={null}
tensorkube job deploy [OPTIONS]
```

**Options:**

* `--name TEXT`: Name of the job to deploy.

* `--gpus INTEGER` \[default: 0]: Number of GPUs needed for an instance of the job.

* `--gpu-type Choice(['V100', 'A10G', 'T4', 'L4', 'L40S', 'A100', 'H100'])`: Type of GPU.

* `--cpu FLOAT` \[default: 100]: Number of CPU millicores. 1000 = 1 CPU

* `--memory FLOAT` \[default: 200]: Amount of RAM in megabytes.

* `--secret TEXT`: Secrets that are required by the job.

* `--max-scale INTEGER` \[default: 3]: Maximum number of jobs to run concurrently.

* `--update BOOLEAN` \[default: False]: Update the job if it already exists.

* `--efs BOOLEAN` \[default: False]: Flag to use EFS for the deployment.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job delete`

Delete a job from Tensorfuse.

This command also de-provisions a job's infrastructure such as queues, buckets and key-value stores.

**Usage:**

```bash theme={null}
tensorkube job delete [OPTIONS]
```

**Options:**

* `--job-name TEXT`: Name of the job to delete.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job queue`

Queue an instance of a deployed job on Tensorfuse.

The payload is a JSON string that is passed to the job instance.

**Usage:**

```bash theme={null}
tensorkube job queue [OPTIONS]
```

**Options:**

* `--job-name TEXT`: Name of the job to delete.

* `--job-id TEXT`: Unique id for the job instance.

* `--payload TEXT`: Payload of the job to delete.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job get`

Get the status of a job instance.

**Usage:**

```bash theme={null}
tensorkube job get [OPTIONS]
```

**Options:**

* `--job-name TEXT`: Name of the job to deploy.

* `--job-id TEXT`: Unique id of the job instance.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job list`

List all jobs or a specific job on Tensorfuse.

1. If no job name is provided, all jobs are listed.
2. If a job name is provided, the status of that job is displayed.

**Usage:**

```bash theme={null}
tensorkube job list [OPTIONS]
```

**Options:**

* `--job-name TEXT`: Name of the job to deploy.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job logs`

Stream the logs for a specific job instance.

This command works only if the job is currently in progress.

**Usage:**

```bash theme={null}
tensorkube job logs [OPTIONS]
```

**Options:**

* `--job-name TEXT`: Name of the job whose logs to display.

* `--help BOOL` \[default: false]: Show this message and exit.

## `tensorkube job oneshot`

IN ALPHA; DO NOT USE. Manage and deploy one-shot jobs on Tensorfuse.

**Usage:**

```bash theme={null}
tensorkube job oneshot COMMAND [ARGS]...
```

**Options:**

* `--help BOOL` \[default: false]: Show this message and exit.

**Commands:**

* `deploy`: No description.