Create Evaluator

curl --request POST \
  --url https://api.fireworks.ai/v1/accounts/{account_id}/evaluatorsV2 \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "evaluator": {
    "displayName": "<string>",
    "description": "<string>",
    "criteria": [
      {
        "type": "TYPE_UNSPECIFIED",
        "name": "<string>",
        "description": "<string>",
        "codeSnippets": {
          "language": "<string>",
          "fileContents": {},
          "entryFile": "<string>",
          "entryFunc": "<string>"
        }
      }
    ],
    "requirements": "<string>",
    "entryPoint": "<string>",
    "commitHash": "<string>",
    "source": {
      "type": "TYPE_UNSPECIFIED",
      "githubRepositoryName": "<string>"
    },
    "defaultDataset": "<string>"
  },
  "evaluatorId": "<string>"
}
'

{
  "name": "<string>",
  "displayName": "<string>",
  "description": "<string>",
  "createTime": "2023-11-07T05:31:56Z",
  "createdBy": "<string>",
  "updateTime": "2023-11-07T05:31:56Z",
  "state": "STATE_UNSPECIFIED",
  "criteria": [
    {
      "type": "TYPE_UNSPECIFIED",
      "name": "<string>",
      "description": "<string>",
      "codeSnippets": {
        "language": "<string>",
        "fileContents": {},
        "entryFile": "<string>",
        "entryFunc": "<string>"
      }
    }
  ],
  "requirements": "<string>",
  "entryPoint": "<string>",
  "status": {
    "code": "OK",
    "message": "<string>"
  },
  "commitHash": "<string>",
  "source": {
    "type": "TYPE_UNSPECIFIED",
    "githubRepositoryName": "<string>"
  },
  "defaultDataset": "<string>"
}

Evaluators

Create Evaluator

Creates a custom evaluator for scoring model outputs. Evaluators use the Eval Protocol to define test cases, run model inference, and score responses. They are used with evaluation jobs and Reinforcement Fine-Tuning (RFT).

Source Code Requirements

Your project should contain:

requirements.txt - Python dependencies for your evaluator
test_*.py - Pytest test file(s) with @evaluation_test decorated functions
Any additional code/modules your evaluator needs

Workflow

Recommended: Use the ep upload CLI command to handle all these steps automatically.

If using the API directly:

Call this endpoint to create the evaluator resource
Package your source directory as a .tar.gz (respecting .gitignore)
Call Get Evaluator Upload Endpoint to get a signed upload URL
PUT the tar.gz file to the signed URL
Call Validate Evaluator Upload to trigger server-side validation
Poll Get Evaluator until ready

Once active, reference the evaluator in Create Evaluation Job or Create Reinforcement Fine-tuning Job.

POST

accounts

{account_id}

evaluatorsV2

Create Evaluator

curl --request POST \
  --url https://api.fireworks.ai/v1/accounts/{account_id}/evaluatorsV2 \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "evaluator": {
    "displayName": "<string>",
    "description": "<string>",
    "criteria": [
      {
        "type": "TYPE_UNSPECIFIED",
        "name": "<string>",
        "description": "<string>",
        "codeSnippets": {
          "language": "<string>",
          "fileContents": {},
          "entryFile": "<string>",
          "entryFunc": "<string>"
        }
      }
    ],
    "requirements": "<string>",
    "entryPoint": "<string>",
    "commitHash": "<string>",
    "source": {
      "type": "TYPE_UNSPECIFIED",
      "githubRepositoryName": "<string>"
    },
    "defaultDataset": "<string>"
  },
  "evaluatorId": "<string>"
}
'

{
  "name": "<string>",
  "displayName": "<string>",
  "description": "<string>",
  "createTime": "2023-11-07T05:31:56Z",
  "createdBy": "<string>",
  "updateTime": "2023-11-07T05:31:56Z",
  "state": "STATE_UNSPECIFIED",
  "criteria": [
    {
      "type": "TYPE_UNSPECIFIED",
      "name": "<string>",
      "description": "<string>",
      "codeSnippets": {
        "language": "<string>",
        "fileContents": {},
        "entryFile": "<string>",
        "entryFunc": "<string>"
      }
    }
  ],
  "requirements": "<string>",
  "entryPoint": "<string>",
  "status": {
    "code": "OK",
    "message": "<string>"
  },
  "commitHash": "<string>",
  "source": {
    "type": "TYPE_UNSPECIFIED",
    "githubRepositoryName": "<string>"
  },
  "defaultDataset": "<string>"
}

Authorizations

Authorization

string

header

required

Bearer authentication using your Fireworks API key. Format: Bearer <API_KEY>

Path Parameters

account_id

string

required

The Account Id

Body

application/json

evaluator

object

required

Show child attributes

evaluatorId

string

Response

200 - application/json

A successful response.

name

string

read-only

displayName

string

description

string

createTime

string<date-time>

read-only

createdBy

string

read-only

updateTime

string<date-time>

read-only

state

enum<string>

default:STATE_UNSPECIFIED

read-only

Available options:

STATE_UNSPECIFIED,

ACTIVE,

BUILDING,

BUILD_FAILED

criteria

Criteria for the evaluator, it should produce a score for the metric (name of criteria) Used for eval3 with UI upload path · object[]

Show child attributes

requirements

string

entryPoint

string

status

Status of the evaluator, used to expose build status to the user · object

read-only

Show child attributes

commitHash

string

source

object

Source information for the evaluator codebase.

Show child attributes

defaultDataset

string

Delete Evaluation Job

List Evaluators

⌘I

Documentation Index

Authorizations

Path Parameters

Body

Response