POST
/
api
/
v1
/
dataset
/
assess
/
upload

​
Parameters

  • evaluation_name: The name of your evaluation

  • dataset_id: (Optional: Required if you are using an existing dataset) The unique identifier for the dataset. Learn more about obtaining a dataset ID.

  • dataset_name: (Optional: Required if you are creating a new dataset) The name for your dataset

  • space_id: The unique identifier for your space. Learn more about obtaining a space ID.

  • file: The file should be in JSON Format

If you are using an existing dataset your file format would be Learn more about obtaining dataset row ids:

[
  {
    "id": "<your dataset_row_id>",
    "judgement": <your_binary_judgement>
  },
  {
    "id": "<your dataset_row_id>",
    "judgement": <your_binary_judgement>
  }
]

If you are creating a new dataset your file format would be:

[
  {
    "conv_prefix": [
      {
        "role": "user",
        "content": "Hey how are you?"
      }
    ],
    "response": {
      "role": "assistant",
      "content": "I'm good, how are you?"
    },
    "judgement": <YOUR JUDGEMENT>
  },
  {
    "conv_prefix": [
      {
        "role": "user",
        "content": "Hey how are you?"
      }
    ],
    "response": {
      "role": "assistant",
      "content": "I'm bad, how are you?"
    },
    "judgement": <YOUR JUDGEMENT>
  }
]

​
Response

Go to evaluation overview page to see your new evaluation

Authorizations

​
Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

multipart/form-data
​
file
file
required
​
evaluation_name
string
required
​
space_id
string
required
​
dataset_id
string
​
dataset_name
string

Response

200
application/json
Successful Response
​
message
string
required
​
data
object | null
Reliability Guardrails API