---
title: "RLHF JSON Format"
slug: "rlhf-json-format"
updated: 2025-04-10T12:00:06Z
published: 2025-04-10T12:00:06Z
canonical: "docs.dataloop.ai/rlhf-json-format"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dataloop.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# RLHF JSON Format

This page describes RLHF JSON, a data format for describing RLHF data in JavaScript Object Notation (JSON).

For more information on RLHF Studio, see the [RLHF Studio](https://docs.dataloop.ai/docs/rlhf).

## The RHLF Studio Item Format

It is built from two layers: a file layer and an annotation layer. The file contains a list of prompts, and the annotations layer contains a list of responses per prompt.

### File or Prompt Layer Format

The following format is a data structure example of a file that contains two prompts. The first prompt contains text and an image, and the second prompt contains text only.

```
{
	"shebang": "dataloop",
	"metadata": {
		"dltype": "prompt"
	},
	"prompts": {
		"prompt1": [
			{
				"mimetype": "application/text",
				"value": "What animal is in this image?"
			},
			{
				"mimetype": "image/jpeg",
				"value": "https://gate.dataloop.ai/api/v1/items/6489600c8d5a1c350e55116a/stream"
			}
		],
		"prompt2": [
			{
				"mimetype": "application/text",
				"value": "What is the eye color of this cat?"
			}
		
		]
	}
}
```

### Annotations or Response Layer Format

Data structure of a response.

```
{
  "id": "64899218661aa36a1744112c",
  "datasetId": "64885c4d71e85c4f03c3758c",
  "url": "https://rc-gate.dataloop.ai/api/v1/items/64899218661aa36a1744112c",
  "dataset": "https://rc-gate.dataloop.ai/api/v1/datasets/64885c4d71e85c4f03c3758c",
  "createdAt": "2023-06-14T10:10:32.863Z",
  "dir": "/image_prompts",
  "filename": "/image_prompts/flyingdog.json",
  "type": "file",
  "hidden": false,
  "metadata": {
    "system": {
      "encoding": "7bit",
      "isBinary": false,
      "mimetype": "application/json",
      "originalname": "flyingdog.json",
      "refs": [],
      "shebang": {
        "dltype": "prompt"
      },
      "size": 161,
      "taskStatusLog": []
    }
  },
  "name": "flyingdog.json",
  "creator": "lior@dataloop.ai",
  "stream": "https://rc-gate.dataloop.ai/api/v1/items/64899218661aa36a1744112c/stream",
  "thumbnail": "https://rc-gate.dataloop.ai/api/v1/items/64899218661aa36a1744112c/thumbnail",
  "annotations": [
    {
      "id": "64899219bf9b191be0acdd2c",
      "datasetId": "64885c4d71e85c4f03c3758c",
      "itemId": "64899218661aa36a1744112c",
      "url": "https://rc-gate.dataloop.ai/api/v1/annotations/64899219bf9b191be0acdd2c",
      "item": "https://rc-gate.dataloop.ai/api/v1/items/64899218661aa36a1744112c",
      "dataset": "https://rc-gate.dataloop.ai/api/v1/datasets/64885c4d71e85c4f03c3758c",
      "type": "binary",
      "label": "q",
      "coordinates": "https://rc-gate.dataloop.ai/api/v1/items/648992159b8b5e823eda9972/stream",
      "metadata": {
        "system": {
          "automated": true,
          "promptId": "first"
        },
        "user": {
          "annotation_type": "prediction",
          "model": {
            "confidence": 0.9,
            "name": "model1"
          },
          "stream": true
        }
      },
      "creator": "lior@dataloop.ai",
      "createdAt": "2023-06-14T10:10:33.092Z",
      "updatedBy": "lior@dataloop.ai",
      "updatedAt": "2023-06-14T10:10:33.092Z",
      "hash": "64885c4d71e85c4f03c3758c_64899218661aa36a1744112c_q_lior@dataloop.ai",
      "source": "sdk"
    }
  ],
  "annotationsCount": 1,
  "annotated": true
}
```

## RLHF JSON Fields Description

| Key Name | Definition | Parent Key |
| --- | --- | --- |
| id | Unique identifier for the item | Root |
| datasetId | Identifier for the dataset containing the item | Root |
| url | API URL to access the item | Root |
| dataset | API URL to access the dataset | Root |
| createdAt | Timestamp of when the item was created | Root |
| dir | Directory path of the item | Root |
| filename | File name of the item | Root |
| type | Type of the item (e.g., file) | Root |
| hidden | Boolean indicating if the item is hidden | Root |
| metadata | Metadata associated with the item | Root |
| system | System-related metadata | metadata |
| encoding | Encoding format of the file | system |
| isBinary | Boolean indicating if the file is binary | system |
| mimetype | MIME type of the file | system |
| originalname | Original name of the file | system |
| refs | References to related items | system |
| shebang | Additional metadata related to file type | system |
| dltype | Type of data stored in the file (e.g., prompt) | shebang |
| size | Size of the file in bytes | system |
| taskStatusLog | Log of task statuses related to the item | system |
| name | Name of the item | Root |
| creator | Email of the user who created the item | Root |
| stream | API URL to stream the item | Root |
| thumbnail | API URL to access the thumbnail | Root |
| annotations | List of annotations associated with the item | Root |
| id | Unique identifier for the annotation | annotations |
| datasetId | Identifier of the dataset associated with the annotation | annotations |
| itemId | Identifier of the item being annotated | annotations |
| url | API URL to access the annotation | annotations |
| item | API URL to access the annotated item | annotations |
| dataset | API URL to access the dataset of the annotation | annotations |
| type | Type of annotation (e.g., binary) | annotations |
| label | Label assigned to the annotation | annotations |
| coordinates | API URL to access coordinates of the annotation | annotations |
| metadata | Metadata related to the annotation | annotations |
| system | System metadata for annotation | metadata |
| automated | Boolean indicating if annotation was automated | system |
| promptId | Identifier for the prompt associated with annotation | system |
| user | User-related metadata | metadata |
| annotation_type | Type of annotation (e.g., prediction) | user |
| model | Model-related metadata | user |
| confidence | Confidence score of the model's prediction | model |
| name | Name of the model | model |
| stream | Boolean indicating if the annotation has a stream | user |
| creator | Email of the user who created the annotation | annotations |
| createdAt | Timestamp when annotation was created | annotations |
| updatedBy | Email of the user who updated the annotation | annotations |
| updatedAt | Timestamp when annotation was last updated | annotations |
| hash | Unique hash identifier for annotation | annotations |
| source | Source of the annotation (e.g., SDK) | annotations |
| annotationsCount | Total count of annotations for the item | Root |
| annotated | Boolean indicating if the item has annotations | Root |

---