Elevate Your Text Generation with cjwbw/dolly's Cognitive Actions

24 Apr 2025
Elevate Your Text Generation with cjwbw/dolly's Cognitive Actions

In the rapidly evolving field of AI-driven content creation, the cjwbw/dolly API provides developers with powerful tools to generate high-quality text. The standout action in this suite is the Generate Text with Dolly-6B, leveraging a fine-tuned version of the GPT-J 6B model on the Alpaca dataset. This action is specifically designed to produce instructional text with enhanced instruction-following capabilities, making it ideal for various text-generation tasks across applications.

Prerequisites

Before you dive into using the Cognitive Actions, ensure you have:

  • An API key for the Cognitive Actions platform to authenticate your requests.
  • Basic knowledge of JSON and API interactions.

Authentication typically involves passing your API key in the headers of your requests, allowing secure access to the Cognitive Actions endpoints.

Cognitive Actions Overview

Generate Text with Dolly-6B

The Generate Text with Dolly-6B action allows you to create high-quality instructional text based on a provided prompt. This action is categorized under text-generation and is perfect for applications needing dynamic content generation.

Input

The input schema for this action requires a JSON object that includes the following properties:

  • prompt (required): The initial text input for the generation model to guide the text generation.
  • topK (optional): Specifies the number of highest probability tokens to consider during generation (default: 50).
  • topP (optional): A cumulative probability cut-off for token selection (default: 1).
  • decoding (optional): The decoding strategy for text generation (default: beam_search).
  • maxLength (optional): The maximum number of tokens to generate (default: 500).
  • numOfBeams (optional): The number of beams for beam search (default: 1).
  • temperature (optional): Controls the randomness of the output (default: 0.75).
  • repetitionPenalty (optional): Adjusts the penalty for word repetition (default: 1).

Here’s an example of the required input JSON:

{
  "topK": 50,
  "topP": 1,
  "prompt": "Write a love letter to Edgar Allan Poe.",
  "decoding": "top_p",
  "maxLength": 500,
  "numOfBeams": 1,
  "temperature": 0.75,
  "repetitionPenalty": 1.2
}

Output

The output from this action is a generated text based on the prompt, reflecting the specifications provided in the input. A typical output example might look like this:

Write a love letter to Edgar Allan Poe.

Dear Edgar, 
I hope this letter finds you well and in good spirits. I wanted to take the time to express my admiration for your work as an author and poet...

The output will include a well-crafted response according to the prompt, with additional metadata about the execution.

Conceptual Usage Example (Python)

Here’s a conceptual Python code snippet demonstrating how to call this action using a hypothetical Cognitive Actions execution endpoint:

import requests
import json

# Replace with your Cognitive Actions API key and endpoint
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"  # Hypothetical endpoint

action_id = "e24aa536-a40e-4549-8771-3933f90603ec"  # Action ID for Generate Text with Dolly-6B

# Construct the input payload based on the action's requirements
payload = {
    "topK": 50,
    "topP": 1,
    "prompt": "Write a love letter to Edgar Allan Poe.",
    "decoding": "top_p",
    "maxLength": 500,
    "numOfBeams": 1,
    "temperature": 0.75,
    "repetitionPenalty": 1.2
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json"
}

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json={"action_id": action_id, "inputs": payload}  # Hypothetical structure
    )
    response.raise_for_status()  # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body: {e.response.text}")

In this snippet, replace YOUR_COGNITIVE_ACTIONS_API_KEY with your actual API key. The action_id is set to the ID for the Generate Text with Dolly-6B action. The input payload is structured according to the requirements of the action.

Conclusion

The cjwbw/dolly API's Cognitive Actions, particularly the Generate Text with Dolly-6B, provide developers with a robust solution for generating instructional text. By leveraging this powerful action, you can enhance your applications with dynamic content generation capabilities. As you explore its features, consider integrating this action into your applications for a seamless user experience in text generation tasks. Happy coding!

← Back to All Articles