Calculate Bleu

POST

/calculate_bleu

curl --request POST \
  --url https://example.com/calculate_bleu \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{ "hypothesis": "example", "reference": "example" }'

Submit a BLEU score calculation task (async).

BLEU (Bilingual Evaluation Understudy) is a precision-based metric that measures how well a machine translation matches a reference translation.

Returns a task_id immediately. Use GET /score/{task_id} to check status and retrieve results.

Authorizations

• HTTPBearer

Request Body^required

Media typeapplication/json

ScoringRequest

object

hypothesis

required

Hypothesis

Machine translation to evaluate

string

>= 1 characters

reference

required

Reference

Reference translation

string

>= 1 characters

Example^generated

{
  "hypothesis": "example",
  "reference": "example"
}

Responses

202

Successful Response

Media typeapplication/json

AsyncScoringSubmission

object

task_id

required

Task Id

Celery task ID for tracking the scoring job

string

status

Status

Initial status of the task

string

default: pending

Example

{
  "status": "pending"
}

422

Validation Error

Media typeapplication/json

HTTPValidationError

object

detail

Detail

Array<object>

ValidationError

object

loc

required

Location

Array

msg

required

Message

string

type

required

Error Type

string

input

Input

ctx

Context

object

Example^generated

{
  "detail": [
    {
      "loc": [
        "example"
      ],
      "msg": "example",
      "type": "example",
      "input": "example",
      "ctx": {}
    }
  ]
}