Skip to main content
POST
/
api
/
v1
/
training
/
datasets
/
{datasetKey}
Start training
curl --request POST \
  --url https://api.mathfi.ai/api/v1/training/datasets/{datasetKey} \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "scalingFactor": 19,
  "performanceThreshold": 0.85
}
'
{
  "trainingJobKey": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "datasetKey": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "status": "PENDING",
  "trainingJobProgressUrl": "<string>"
}

Authorizations

​
Authorization
string
header
required

Obtain a token from POST /api/login. Valid for 1 hour.

Path Parameters

​
datasetKey
string<uuid>
required

Body

application/json
​
scalingFactor
integer<int32>
required

Controls the internal learning rate. A starting value of 10–50 works for most datasets. Adjust based on training performance progression.

Required range: 0 <= x <= 499
Example:

19

​
performanceThreshold
number<double>
required

Target balanced-class accuracy (0–1). Training stops when any algorithm reaches this value. Lower values complete faster. Typical starting point: 0.80.

Required range: 0 <= x <= 1
Example:

0.85

Response

Training job created successfully

​
trainingJobKey
string<uuid>
required
​
datasetKey
string<uuid>
required
​
status
enum<string>
required

Status of a training job:

StatusDescription
PENDINGQueued, not yet started
RUNNINGAlgorithms competing in parallel
COMPLETEDTarget performance reached. Champion model created.
TIMED_OUTDid not reach target performance within the time limit
NOT_COMPLETEDStalled — no progress detected for a sustained period
CANCELLEDCancelled by the user
FAILEDFailed due to an internal error
Available options:
PENDING,
RUNNING,
COMPLETED,
TIMED_OUT,
NOT_COMPLETED,
CANCELLED,
FAILED
​
trainingJobProgressUrl
string

URL to poll for training progress