# Start Training

Start model training. Training duration is typically 2-4 hours.
The associated dataset must have a status of 'completed' before training can begin.

Defaults by Training Version:
If parameters are not provided, default values are automatically selected based on the model's training_version:

* FIBO (fibo):
    * rank: 128
    * learning_rate: 0.0001
    * lr_scheduler: "constant_with_warmup"
    * total_training_steps: 750
    * checkpoint_interval: 500

* Legacy (max, light, 3.2, 2.3):
    * rank: 128
    * learning_rate: 0.0001
    * lr_scheduler: "constant"
    * total_training_steps: 1500
    * checkpoint_interval: 500
    
Constraints:
* Advanced training parameters are only supported when the model's training_mode is set to 'expert'.

Endpoint: POST /tailored-gen/models/{model_id}/start_training

## Path parameters:

  - `model_id` (integer, required)
    Model ID

## Header parameters:

  - `api_token` (string, required)

## Request fields (application/json):

  - `lr_scheduler` (string)
    The learning rate scheduler used to adjust the learning rate throughout the training process.
* Default for FIBO: constant_with_warmup
* Default for Legacy: constant
    Enum: "linear", "cosine", "cosine_with_restarts", "polynomial", "constant", "constant_with_warmup"

  - `learning_rate` (number)
    The initial learning rate. Default is 0.0001.

  - `total_training_steps` (integer)
    The total number of training optimization steps.
* Default for FIBO: 750
* Default for Legacy: 1500

  - `rank` (integer)
    The rank (latent dimension). Default is 128.

  - `checkpoint_interval` (integer)
    The interval (in steps) at which model checkpoints are saved. Default is 500.

  - `lr_warmup_steps` (integer)
    Number of warmup steps for the learning rate scheduler.
* Relevant only when lr_scheduler is set to constant_with_warmup.
* Applies to all training versions.

## Response 200 fields (application/json):

  - `id` (integer)
    Unique identifier for the model.

  - `name` (string)
    Name of the model.

  - `description` (string)
    Description of the model.

  - `status` (string)
    Status of the model.
    Enum: "InProgress"

  - `dataset_id` (integer)
    ID of the dataset used for training.

  - `training_version` (string)
    The training version of the model.

  - `training_mode` (string)
    The training mode of the model.

  - `generation_prefix` (string,null)
    Text automatically prepended to prompts.
* For fibo models: Null.
* For legacy models: String matching the dataset's caption_prefix.

  - `project_id` (integer)
    ID of the project this model belongs to.

  - `created_at` (string)
    Timestamp when the model was created.

  - `updated_at` (string)
    Timestamp when the model was last updated.

  - `custom_metadata` (string)
    Custom metadata associated with the model.

  - `active_model_version` (string)
    The active version of the model.

  - `training_parameters` (object)
    The training parameters used for the model.

  - `training_parameters.learning_rate` (number)

  - `training_parameters.lr_scheduler` (string)

  - `training_parameters.rank` (integer)

  - `training_parameters.total_training_steps` (integer)

  - `training_parameters.checkpoint_interval` (integer)

  - `training_parameters.lr_warmup_steps` (integer,null)
    Number of warmup steps. Returns null if lr_scheduler is NOT constant_with_warmup.

  - `checkpoint_steps` (array)
    A list of checkpoint steps for the model.


## Response 401 fields

## Response 404 fields

## Response 409 fields

## Response 412 fields

## Response 429 fields

## Response 500 fields