Unlocking Code Generation with WizardCoder 34B Cognitive Actions

23 Apr 2025
Unlocking Code Generation with WizardCoder 34B Cognitive Actions

In the world of programming, automating code generation can significantly boost productivity and efficiency. The WizardCoder 34B, part of the rhamnett/wizardcoder-34b-v1.0 specification, is a powerful model that excels at generating high-quality code. It's a variant of Code Llama and has been shown to outperform GPT-4 on Human Eval benchmarks. With its customizable parameters, developers can fine-tune the output to meet their specific needs. In this article, we'll explore how to leverage the Generate Code Using WizardCoder 34B action to enhance your applications.

Prerequisites

Before diving into the usage of the WizardCoder 34B Cognitive Actions, ensure you have the following:

  • An API key for the Cognitive Actions platform.
  • Basic knowledge of making API requests.
  • An environment set up for running Python scripts.

To authenticate your requests, you'll typically include your API key in the headers of your HTTP requests, allowing you access to the action functionalities.

Cognitive Actions Overview

Generate Code Using WizardCoder 34B

The Generate Code Using WizardCoder 34B action is designed for generating code snippets based on a given prompt. It allows users to customize various parameters such as temperature, max length, and repetition penalty to control the output quality and style.

Input

The input for this action requires several fields, primarily the prompt, which is the instruction for the model. Here’s a breakdown of the input schema:

  • prompt (required): A string that contains the instruction for the code generation.
    Example: "Write a javascript function that calculates euclidean distance between two coordinates of any dimension"
  • maxLength (optional): An integer defining the maximum number of tokens the model can generate. Default is 256.
  • temperature (optional): A number that controls the randomness of the output. Default is 0.75.
  • topP (optional): A number that sets the probability threshold for token sampling. Default is 1.
  • numberOfSequences (optional): An integer specifying the number of distinct output sequences to generate. Default is 1.
  • repetitionPenalty (optional): A number that applies a penalty to repeated words in the output. Default is 1.1.

Here’s an example input JSON payload for this action:

{
  "topP": 1,
  "prompt": "Write a javascript function that calculates euclidean distance between two coordinates of any dimension",
  "maxLength": 256,
  "temperature": 0.75,
  "numberOfSequences": 1,
  "repetitionPenalty": 1.1
}

Output

The output from this action typically includes the generated code snippet based on the provided prompt. For example, using the above input, the output could look like this:

[
  "Write a javascript function that calculates euclidean distance between two coordinates of any dimension. The input parameters are the two coordinate arrays and the dimensions of the coordinates:\r\n\r\n```javascript\r\nfunction calculateDistance(coord1, coord2) {\r\n  let dimension = coord1.length;\r\n  let distance = 0;\r\n  for (let i = 0; i < dimension; i++) {\r\n    distance += Math.pow(coord1[i] - coord2[i], 2);\r\n  }\r\n  return Math.sqrt(distance);\r\n}\r\n```\r\n\r\nExample usage:\r\n\r\n```javascript\r\nlet coord1 = [0, 0];\r\nlet coord2 = [3, 4];\r\nconsole.log(calculateDistance(coord1, coord2)); // Output: 5\r\n\r\nlet coord3 = [1, 2, 3];\r\nlet coord4 = [4, 6, 9];\r\nconsole.log(calculateDistance(coord3, coord4)); // Output: 7.071067811865475"
]

Conceptual Usage Example (Python)

Here’s how you might call this action using a hypothetical Cognitive Actions endpoint in Python:

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 = "e0ad3953-44cb-4ef1-ab74-161b52a13cc2"  # Action ID for Generate Code Using WizardCoder 34B

# Construct the input payload based on the action's requirements
payload = {
    "topP": 1,
    "prompt": "Write a javascript function that calculates euclidean distance between two coordinates of any dimension",
    "maxLength": 256,
    "temperature": 0.75,
    "numberOfSequences": 1,
    "repetitionPenalty": 1.1
}

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 example, the code constructs the input payload based on the required parameters and sends a POST request to the action execution endpoint. The action ID and the input payload are highlighted for clarity.

Conclusion

The Generate Code Using WizardCoder 34B action offers developers a robust tool for automating code generation tasks, enhancing productivity and creativity in software development. By customizing parameters such as temperature and max length, you can tailor the output to suit your specific coding needs. Now that you have the knowledge to integrate this action, consider exploring various prompts and configurations to see how WizardCoder can transform your coding workflows. Happy coding!

← Back to All Articles