Dataset and Dataset Items - Zenbase AI

Datasets

A Dataset accepts the following parameters:

name (str): The name of the dataset
description (str): A description of what the dataset is
embeddig_api_key (str): the API key for generating embeddings
embeddig type (str): For now, only supports OPENAI
model_name (str): The name of the model to use for generating embeddings

Embedding type and model name are only required if you want to use online optimization, or you want to use the search_similar endpoint. Example of how to create a dataset:

import requests
import json

BASE_URL = "https://orch.zenbase.ai/api"
API_KEY = "YOUR ZENBASE API KEY"

def api_call(method, endpoint, data=None):
    url = f"{BASE_URL}/{endpoint}"
    headers = {
        "Content-Type": "application/json",
        "Authorization": f"Api-Key {API_KEY}"
    }
    response = requests.request(method, url, headers=headers, data=json.dumps(data) if data else None)
    return response

dataset_data = {
    "name": "my-dataset",
    "description": "This is my dataset",
    "embedding_api_key": "MY API KEY",
    "embedding_type": "OPENAI",
    "model_name": "text-embedding-ada-002",
}
dataset = api_call("POST", "datasets/", dataset_data)
dataset_id = dataset.json()['id']

Dataset Items

A Dataset Item accepts the following parameters:

dataset (int): The id of the dataset this item belongs to
inputs (dict): The input data for the dataset item
outputs (dict): The output data for the dataset item

Example of how to create a dataset item:

import requests
import json

BASE_URL = "https://orch.zenbase.ai/api"
API_KEY = "YOUR ZENBASE API KEY"

def api_call(method, endpoint, data=None):
    url = f"{BASE_URL}/{endpoint}"
    headers = {
        "Content-Type": "application/json",
        "Authorization": f"Api-Key {API_KEY}"
    }
    response = requests.request(method, url, headers=headers, data=json.dumps(data) if data else None)
    return response


dataset_item_data = {
    "dataset": 1, # The id of the dataset this item belongs to
    "inputs": {"text": "This is my input"},
    "outputs": {"sentiment": "positive"},
}
response = api_call("POST", "dataset-items/", dataset_item_data)