> ## Documentation Index > Fetch the complete documentation index at: https://nixtla.io/docs/llms.txt > Use this file to discover all available pages before exploring further. # Submit an async finetune job > Queues a finetune job and returns immediately with its `job_id`. The job runs in a sandbox; poll `GET /v2/finetune/jobs/{job_id}` for its state and result. Accepts the same body as the synchronous endpoint plus an optional `job_options`. ## OpenAPI ````yaml /openapi.json post /v2/finetune/async openapi: 3.1.0 info: title: Nixtla Forecast API description: >- API for TimeGPT forecast. Just send your data as json and get results. We do the heavy lifting. version: 0.4.2 servers: - url: https://api.nixtla.io security: [] paths: /v2/finetune/async: post: tags: - async jobs summary: Submit an async finetune job description: >- Queues a finetune job and returns immediately with its `job_id`. The job runs in a sandbox; poll `GET /v2/finetune/jobs/{job_id}` for its state and result. Accepts the same body as the synchronous endpoint plus an optional `job_options`. operationId: submit_async_finetune_v2_finetune_async_post requestBody: content: application/json: schema: $ref: '#/components/schemas/FinetuneAsyncRequest' examples: - finetune_steps: 10 freq: MS model: timegpt-1 series: sizes: - 36 'y': - 0 - 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 - 19 - 20 - 21 - 22 - 23 - 24 - 25 - 26 - 27 - 28 - 29 - 30 - 31 - 32 - 33 - 34 - 35 required: true responses: '202': description: Successful Response content: application/json: schema: $ref: '#/components/schemas/AsyncJobSubmitResponse' '422': description: Validation Error content: application/json: schema: $ref: '#/components/schemas/HTTPValidationError' security: - HTTPBearer: [] components: schemas: FinetuneAsyncRequest: properties: series: $ref: '#/components/schemas/SeriesWithFutureExogenous' freq: type: string title: Freq description: >- The frequency of the data represented as a string. 'D' for daily, 'M' for monthly, 'H' for hourly, and 'W' for weekly frequencies are available. model: type: string title: Model description: >- Model to use as a string. Common options are (but not restricted to) `timegpt-1` and `timegpt-1-long-horizon`. Full options vary by different users. Contact support@nixtla.io for more information. We recommend using `timegpt-1-long-horizon` for forecasting if you want to predict more than one seasonal period given the frequency of your data. default: timegpt-1 finetune_steps: type: integer minimum: 0 title: Finetune Steps description: >- The number of tuning steps used to train the large time model on the data. Set this value to 0 for zero-shot inference, i.e., to make predictions without any further model tuning. default: 10 finetune_loss: type: string enum: - default - mae - mse - rmse - mape - smape - poisson title: Finetune Loss description: >- The loss used to train the large time model on the data. Select from ['default', 'mae', 'mse', 'rmse', 'mape', 'smape', 'poisson']. It will only be used if finetune_steps is larger than 0. Default is a robust loss function that is less sensitive to outliers. default: default finetune_depth: type: integer enum: - 1 - 2 - 3 - 4 - 5 title: Finetune Depth description: >- The depth of the finetuning. Uses a scale from 1 to 5, where 1 means little finetuning, and 5 means that the entire model is finetuned. Note that this parameter is only effective for timegpt-1 and timegpt-1-long-horizon models; it has no effect on the other models. By default, the value is set to 1. default: 1 output_model_id: anyOf: - type: string pattern: ^[a-zA-Z0-9\-_]{1,36}$ - type: 'null' title: Output Model Id description: ID to assign to the finetuned model finetuned_model_id: anyOf: - type: string pattern: ^[a-zA-Z0-9\-_]{1,36}$ - type: 'null' title: Finetuned Model Id description: ID of previously finetuned model hist_exog: anyOf: - items: type: integer minimum: 0 type: array - type: 'null' title: Hist Exog description: Zero-based indices of the exogenous features to treat as historical. multivariate: type: boolean title: Multivariate description: >- Compute multivariate predictions across a batch of multiple time series. Requires all time series with overlapping dates. Note that this is only effective for timegpt-2.1 model and it has no effect on the other models. By default, the value is set to False. default: false model_parameters: anyOf: - additionalProperties: true type: object - type: 'null' title: Model Parameters description: >- Optional dictionary of parameters to customize the behavior of the large time model. job_options: anyOf: - $ref: '#/components/schemas/AsyncJobOptions' - type: 'null' type: object required: - series - freq title: FinetuneAsyncRequest AsyncJobSubmitResponse: properties: job_id: type: string title: Job Id description: >- Identifier for the accepted job. Prefixed per task (e.g. `fc-` for forecast). examples: - fc-4f2a1c9e8b7d4a6f9c3e1b5d7a9f2c4e type: object required: - job_id title: AsyncJobSubmitResponse description: >- The 202 every async submit returns. Shared by all tasks — they differ in request, not reply. HTTPValidationError: properties: detail: items: $ref: '#/components/schemas/ValidationError' type: array title: Detail type: object title: HTTPValidationError SeriesWithFutureExogenous: properties: X_future: anyOf: - items: items: anyOf: - type: number - type: string type: array type: array - type: 'null' title: X Future description: >- Future values of the exogenous features. Each feature must be a list of size number of series times the forecast horizon (h). X: anyOf: - items: items: anyOf: - type: number - type: string type: array type: array - type: 'null' title: X description: >- Historic values of the exogenous features. Each feature must be a list of the same size as the target (y). categorical_exog: anyOf: - items: type: integer minimum: 0 type: array - type: 'null' title: Categorical Exog description: >- Zero-based indices of the columns in X that are categorical features. 'y': items: type: number type: array title: 'Y' description: Historic values of the target. sizes: items: type: integer type: array title: Sizes description: Sizes of the individual series. start_datetime: anyOf: - items: type: string type: array - type: 'null' title: Start Datetime description: >- Starting timestamp of each individual series, as ISO 8601 strings (for example '2021-01-01' or '2021-01-01T09:30:00'). One entry per series, so it must have the same length as `sizes`. Together with `sizes` and `freq` this recovers the timestamps of every series. type: object required: - 'y' - sizes title: SeriesWithFutureExogenous AsyncJobOptions: properties: timeout_seconds: anyOf: - type: integer exclusiveMinimum: 0 - type: 'null' title: Timeout Seconds additionalProperties: false type: object title: AsyncJobOptions ValidationError: properties: loc: items: anyOf: - type: string - type: integer type: array title: Location msg: type: string title: Message type: type: string title: Error Type input: title: Input ctx: type: object title: Context type: object required: - loc - msg - type title: ValidationError securitySchemes: HTTPBearer: type: http description: HTTPBearer scheme: bearer ````