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

# toloka batch-inference run

> Quote a Batch Inference Run, then confirm it to start.

## Usage

```bash theme={null}
toloka batch-inference run [options]
```

## Options

**--dataset \<id>** `string` *required*

Dataset whose chat messages are the prompts.

**--filter \<field=value>** `string`

Only rows whose field equals the value.

**--slice \<id>** `string`

Only rows of this ready Dataset Slice, with the prompts it froze.

**--parts \<names>** `string`

Only rows in these split parts, comma-separated, for example test.

**--split-column \<name-or-id>** `string`

Split column that `--parts` names, when the dataset has more than one.

**--base-model \<id>** `string`

Catalog Base Model that answers; pass exactly one of `--base-model` or `--model-version`.

**--model-version \<id>** `string`

Model Version that answers: mdv.\<id> or Model\@name; pass exactly one of `--base-model` or `--model-version`.

**--output-field \<name>** `string`

Text field on the dataset that receives the answers; pass exactly one of `--output-field` or `--output-dataset`.

**--output-dataset \<name>** `string`

New dataset that receives the prompts and answers; pass exactly one of `--output-field` or `--output-dataset`.

**--max-output-tokens \<count>** `integer` *required*

Token limit per answer, at most 8192.

**--temperature \<value>** `number`

Sampling temperature from 0 to 2 (default 0).

**--top-p \<value>** `number`

Nucleus sampling threshold above 0 and at most 1.

**--stop \<text>** `string` (default: \[])

Stop sequence; repeat for up to four.

Repeat this option to supply multiple values.

**--seed \<number>** `integer`

Sampling seed.

**--samples-per-row \<count>** `integer`

Answers per row, at most 16; above 1 needs `--output-dataset` and a temperature above 0.

**--idempotency-key \<key>** `string`

Reuse the run of an earlier identical request.

**--yes** `boolean` (default: false)

Confirm the quoted maximum charge and start right away.

**--json** `boolean` (default: false)

Print JSON.

## Inherited options

**-v, --version** `boolean`

Output the current version.

**--source-root \<directory>** `string`

Project root; otherwise search the working directory and parents; init and report use only the working directory.

**--api-key \<key>** `string`

API key override; login stores the key it verifies.

Environment fallback: [`TOLOKA_SCOPED_TOKEN`](https://docs.platform.toloka.ai/docs/cli/reference#environment-variables), [`TOLOKA_API_KEY`](https://docs.platform.toloka.ai/docs/cli/reference#environment-variables).

**--base-url \<url>** `string`

Platform base URL override; login saves the selected URL.

Environment fallback: [`TOLOKA_BASE_URL`](https://docs.platform.toloka.ai/docs/cli/reference#environment-variables).

**--project-id \<id>** `string`

Target project; overrides toloka.json.

Environment fallback: [`TOLOKA_PROJECT_ID`](https://docs.platform.toloka.ai/docs/cli/reference#environment-variables).

See the [CLI guide](https://docs.platform.toloka.ai/docs/cli.md#train) for the workflow, and [environment variables and exit codes](https://docs.platform.toloka.ai/docs/cli/reference.md#environment-variables) for shared behavior.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.