Classifiers based on YandexGPT

Yandex AI Studio allows you to classify text requests provided in prompts. Classification in YandexGPT-based models is implemented in the AI Studio Text Classification API.

There are three types of classification available in AI Studio:

• Binary classification assigns a request to one of two classes, e.g., spam or non-spam.
• Multi-class classification puts a request into one (and only one) of more than two classes. For example, a computer CPU can belong to one generation only.
• Multi-label classification allows you to assign a request to multiple different non-mutually exclusive classes at the same time. For example, a single social media post may have multiple hashtags at the same time.

Classification models are only available in synchronous mode.

AI Studio provides two types of classifiers:

• Prompt-based classifiers based on YandexGPT Lite and YandexGPT Pro: Binary and multi-class classification.
• Fine-tunable classifiers based on YandexGPT Lite: All supported classification types.

To use the Yandex AI Studio classifier models, you need the ai.languageModels.user role or higher for the folder.

Prompt-based classifiersPrompt-based classifiers

AI Studio prompt-based classifiers support binary and multi-class classification, require no model tuning, and are prompt-controlled. The fewShotClassify Text Classification API method allows using these two prompt-based classifiers: Zero-shot and few-shot. You can provide between 2 and 20 classes to the fewShotClassify method.

Zero-shot classifierZero-shot classifier

The zero-shot classifier allows you to perform binary and multi-class classification by providing only the model ID, task description, request text, and an array of class names in the request body.

Request body format for the zero-shot classifier:

{
  "modelUri": "string",
  "taskDescription": "string",
  "labels": [
    "string",
    "string",
    ...
    "string"
  ],
  "text": "string"
}

Where:

• modelUri: ID of the model that will be used to classify the message. This parameter contains the Yandex Cloud folder ID.

• taskDescription: Text description of the task for the classifier.

• labels: Array of classes.

  Give meaningful names to classes in labels: this is essential for correct classification results. For example, use chemistry and physics rather than chm and phs for class names.

• text: Message text.

Use the https://llm.api.cloud.yandex.net/foundationModels/v1/fewShotTextClassification endpoint for requests to zero-shot classifiers.

Few-shot classifierFew-shot classifier

The few-shot classifier enables binary and multi-class classification by providing the model with an array of sample requests for the classes specified in the labels field. You need to list these sample requests in the samples field of the request body to get more accurate results from the classifier.

Request body format for the few-shot classifier:

{
  "modelUri": "string",
  "taskDescription": "string",
  "labels": [
    "string",
    "string",
    ...
    "string"
  ],
  "text": "string",
  "samples": [
    {
      "text": "string",
      "label": "string"
    },
    {
      "text": "string",
      "label": "string"
    },
    ...
    {
      "text": "string",
      "label": "string"
    }
  ]
}

Where:

• modelUri: ID of the model that will be used to classify the message. This parameter contains the Yandex Cloud folder ID.

• taskDescription: Text description of the task for the classifier.

• labels: Array of classes.

  Give meaningful names to classes in labels: this is essential for correct classification results. For example, use chemistry and physics rather than chm and phs for class names.

• text: Message text.

• samples: Array with examples of prompts for the classes specified in the labels field. Examples of prompts are provided as objects, each containing an example of a text query and the class to which such query should belong.

Use the https://llm.api.cloud.yandex.net/foundationModels/v1/fewShotTextClassification endpoint for requests to few-shot classifiers.

Trainable classifiersTrainable classifiers

If not satisfied with the output quality of the zero-shot and few-shot classifiers, tune your own one based on YandexGPT Lite. Trainable classifiers can be trained to offer all supported classification types.

To run a request to a fine-tuned classifier, use the classify text classification API method. If you do so, you only need to provide the model ID and the request text to the model. The names of the classes between which the model will be distributing requests must be specified during model tuning and are not provided in the request.

Request body format for a fine-tuned classifier:

{
  "modelUri": "string",
  "text": "string"
}

Where:

• modelUri: ID of the model that will be used to classify the message. This parameter contains the Yandex Cloud folder ID and the ID of the fine-tuned classifier.
• text: Message text. The total number of tokens per request must not exceed 8,000.

Use the https://llm.api.cloud.yandex.net:443/foundationModels/v1/textClassification endpoint for requests to trainable classifiers.

The names of the classes between which the model will be distributing requests must be specified during model tuning and are not provided in the request.

Response formatResponse format

All AI Studio classifier types return the result in the following format:

{
  "predictions": [
    {
      "label": "string",
      "confidence": "number",
    },
    {
      "label": "string",
      "confidence": "number",
    },
    ...
    {
      "label": "string",
      "confidence": "number",
    }
  ],
  "modelVersion": "string"
}

Where:

• label: Class name.

• confidence: Probability of assigning the request text to this class.

  In multi-class classification, the sum of the confidence values for all classes is always 1.

  In multi-label classification, the confidence value for each class is calculated independently (the sum of the values is not equal to 1).

Use casesUse cases

• Using fine-tuned classifiers based on YandexGPT
• Using prompt-based classifiers based on YandexGPT