Create a training job

POST

https://api.fastino.ai

training-jobs

curl -X POST https://api.fastino.ai/v1/training-jobs \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "my-model-name",
    "base_model": "fastino/gliner2-multi-v1",
    "datasets": [{"name": "my-ready-dataset"}],
    "training_type": "lora",
    "nr_epochs": 5,
    "learning_rate": 5e-5,
    "validation_data_percentage": 0.2
  }'

{
  "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
  "model_name": "my-model-name",
  "base_model": "fastino/gliner2-multi-v1",
  "status": "requested",
  "training_type": "lora",
  "nr_epochs": 5,
  "learning_rate": 5e-5,
  "validation_data_percentage": 0.2
}

POST /v1/training-jobs Creates an asynchronous training job and returns its job record. This endpoint is limited to 20 requests per minute per user.

Request

string

required

Your Fastino API key.

string

required

Display name for the trained model. Length: 1–100 characters.

string

required

A trainable model ID from GET /v1/base-models?supports_training=true, or a supported checkpoint UUID.

object[]

required

One or more dataset references. Each item requires name and may include version; omitted versions resolve to the latest version.

string

default:"lora"

lora or full. The selected base model must support the requested type.

number

default:"0.2"

Fraction held out for validation, from 0 to 1.

integer

default:"100"

Maximum epoch count. Must be at least 1; early stopping can finish sooner.

number

Positive peak learning rate. Omit it to use the selected model’s training recipe.

integer

default:"4"

Per-device batch size. Prefer omitting it so the training service can apply or safely clamp to the model-specific default.

integer

Optional reproducibility seed from 0 through 2147483647. Set provider_name to modal when using it. A seed reduces one source of variation but does not guarantee bit-identical GPU runs.

string

Project UUID. When omitted, Fastino associates the job with the caller’s default project.

Show Additional training parameters

integer

default:"100"

Save a checkpoint every N steps.

boolean

default:"false"

Persist a structured training profile artifact.

string

Optional Weights & Biases API key. Treat it as a secret and never put it in source control, logs, or examples.

integer

LoRA rank. Omit it to use the model recipe.

integer

LoRA alpha. Omit it to use the model recipe.

number

LoRA dropout. Omit it to use the model recipe.

boolean

Pack short examples for compatible decoder LoRA jobs.

boolean

default:"false"

Decoder loss-masking option. Unsupported combinations are rejected.

number

Warmup fraction from 0 to 1. warmup_steps takes precedence.

integer

Positive absolute warmup-step count.

string

default:"cosine"

Learning-rate schedule: constant, linear, or cosine.

number

default:"0.01"

Non-negative AdamW weight decay.

integer

default:"3"

Validation epochs without improvement before stopping. Set 0 to disable.

number

default:"0.0001"

Minimum validation-loss improvement that counts as progress.

string

Pin a compatible training provider. Omit it for automatic selection.

string

Canonical system prompt for compatible decoder training.

number

GLiNER encoder learning rate. Falls back to learning_rate.

number

GLiNER task-head learning rate. Falls back to learning_rate.

integer

Positive mini-batch accumulation count.

boolean

Enable GLiNER dataset auto-sizing.

integer

GLiNER auto-sizing lower bound.

integer

GLiNER auto-sizing upper bound.

integer

GLiNER auto-sizing scale per label.

integer

Minimum GLiNER optimizer steps.

string

default:"sft"

sft, grpo, or dpo. GRPO requires rl_config.reward_type. DPO datasets require prompt, chosen, and rejected columns.

object

Algorithm-specific settings. Shared keys include max_steps and logging_steps. GRPO supports reward_type, kl_beta, group_size, sampling_temperature, and max_completion_length. DPO supports dpo_beta and loss_type. The llm_as_judge GRPO reward also supports llm_judge_model, llm_judge_rubric, llm_judge_score_scale, llm_judge_timeout_s, llm_judge_max_concurrent, llm_judge_max_retries, and llm_judge_retry_backoff_s.

curl -X POST https://api.fastino.ai/v1/training-jobs \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "my-model-name",
    "base_model": "fastino/gliner2-multi-v1",
    "datasets": [{"name": "my-ready-dataset"}],
    "training_type": "lora",
    "nr_epochs": 5,
    "learning_rate": 5e-5,
    "validation_data_percentage": 0.2
  }'

Response

Returns 200 with the full training-job record, including a UUID id and initial status.

{
  "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
  "model_name": "my-model-name",
  "base_model": "fastino/gliner2-multi-v1",
  "status": "requested",
  "training_type": "lora",
  "nr_epochs": 5,
  "learning_rate": 5e-5,
  "validation_data_percentage": 0.2
}

Save the id from this response. You will use it to poll status, retrieve metrics and artifacts, and run inference against the trained model. A malformed body returns 422. A valid but unavailable model, dataset, training type, or provider combination returns a request-specific 4xx error.

⌘I

curl -X POST https://api.fastino.ai/v1/training-jobs \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "my-model-name",
    "base_model": "fastino/gliner2-multi-v1",
    "datasets": [{"name": "my-ready-dataset"}],
    "training_type": "lora",
    "nr_epochs": 5,
    "learning_rate": 5e-5,
    "validation_data_percentage": 0.2
  }'

{
  "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6",
  "model_name": "my-model-name",
  "base_model": "fastino/gliner2-multi-v1",
  "status": "requested",
  "training_type": "lora",
  "nr_epochs": 5,
  "learning_rate": 5e-5,
  "validation_data_percentage": 0.2
}

Get Started

Core Concepts

Inference API

Training API

Account

Request

Response

​Request

​Response

Request

Response