Optimization

Optimization run endpoints.

Create an optimization run

post

Create a new optimization run for a container using specified base models and evaluation criteria.

Authorizations

ApiKeyAuth

AuthorizationstringRequired

API key in Authorization header using Bearer .

Body

application/json

containerstring · min: 1Required

Container id or label.

Example: file_abc123

base_modelsstring[]Required

Base model ids or labels.

evalsstring[]Required

Evaluator ids or labels.

data_sourceany ofOptional

Optional data source override.

anyOptional

or

or

or

nullOptional

metadataany ofOptional

Optional metadata.

or

nullOptional

environmentany ofOptional

Execution environment name.

Default: main

stringOptional

or

nullOptional

Responses

201

Successful Response

application/json

created_atstringRequired

finished_atany ofOptional

stringOptional

or

nullOptional

error_atany ofOptional

stringOptional

or

nullOptional

statusstringRequired

errorany ofOptional

anyOptional

or

nullOptional

objectconst: optimization.runOptional

Object type.

Default: optimization.run

idstringRequired

Optimization run id.

container_idstringRequired

Container id.

base_modelsstring[]Required

Base model ids.

datasetany ofOptional

Dataset id.

stringOptional

or

nullOptional

evaluatorsstring[]Required

Evaluator ids.

metadataany ofOptional

Optional metadata.

or

nullOptional

metricsany ofOptional

Optimization metrics.

or

nullOptional

spendany ofOptional

Estimated spend.

numberOptional

or

nullOptional

best_checkpointany ofOptional

Best checkpoint id.

stringOptional

or

nullOptional

400

Bad Request

application/json

401

Unauthorized

application/json

403

Forbidden

application/json

404

Not Found

application/json

409

Conflict

application/json

422

Validation Error

application/json

429

Too Many Requests

application/json

500

Internal Server Error

application/json

501

Not Implemented

application/json

503

Upstream Unavailable

application/json

post

/v1/optimization/runs

HTTPcURLJavaScriptPython

POST /v1/optimization/runs HTTP/1.1
Host: platform.maniac.ai
Authorization: Bearer YOUR_SECRET_TOKEN
Content-Type: application/json
Accept: */*
Content-Length: 778

{
  "container": "file_abc123",
  "base_models": [
    "file_abc123"
  ],
  "evals": [
    "file_abc123"
  ],
  "methods": [
    {
      "type": "text",
      "hyperparameters": {
        "max_steps": 1,
        "lora_targets": [
          "text"
        ],
        "lora_dim": [
          1
        ],
        "lora_alpha": [
          1
        ],
        "per_device_train_batch_size": [
          1
        ],
        "gradient_accumulation_steps": [
          1
        ],
        "warmup_ratio": [
          1
        ],
        "optim": [
          "text"
        ],
        "weight_decay": [
          1
        ],
        "lr_scheduler_type": [
          "text"
        ],
        "learning_rate": [
          1
        ],
        "seed": [
          1
        ],
        "mode": [
          "autoregression"
        ],
        "num_eval_checkpoints": 7,
        "train_range": "500:20000",
        "eval_range": "0:500",
        "top_c": 10,
        "filter": "text",
        "max_seq_len": 2048,
        "max_gen_tokens": 1,
        "target_key": "relevance",
        "classification_labels": [
          "text"
        ],
        "ANY_ADDITIONAL_PROPERTY": "anything"
      }
    }
  ],
  "data_source": {
    "range": "0:100",
    "type": "text",
    "file": "file_abc123"
  },
  "metadata": {
    "ANY_ADDITIONAL_PROPERTY": "anything"
  },
  "environment": "main"
}

{
  "created_at": "text",
  "finished_at": "text",
  "error_at": "text",
  "status": "text",
  "error": null,
  "object": "optimization.run",
  "id": "text",
  "container_id": "text",
  "base_models": [
    "text"
  ],
  "methods": [
    {
      "type": "text",
      "hyperparameters": {
        "max_steps": 1,
        "lora_targets": [
          "text"
        ],
        "lora_dim": [
          1
        ],
        "lora_alpha": [
          1
        ],
        "per_device_train_batch_size": [
          1
        ],
        "gradient_accumulation_steps": [
          1
        ],
        "warmup_ratio": [
          1
        ],
        "optim": [
          "text"
        ],
        "weight_decay": [
          1
        ],
        "lr_scheduler_type": [
          "text"
        ],
        "learning_rate": [
          1
        ],
        "seed": [
          1
        ],
        "mode": [
          "autoregression"
        ],
        "num_eval_checkpoints": 7,
        "train_range": "500:20000",
        "eval_range": "0:500",
        "top_c": 10,
        "filter": "text",
        "max_seq_len": 2048,
        "max_gen_tokens": 1,
        "target_key": "relevance",
        "classification_labels": [
          "text"
        ],
        "ANY_ADDITIONAL_PROPERTY": "anything"
      }
    }
  ],
  "dataset": "text",
  "evaluators": [
    "text"
  ],
  "metadata": {
    "ANY_ADDITIONAL_PROPERTY": "anything"
  },
  "metrics": {
    "ANY_ADDITIONAL_PROPERTY": "anything"
  },
  "spend": 1,
  "best_checkpoint": "text"
}