--- title: MiniGpt-4 API Usage Guide createdAt: Thu Jul 18 2024 06:12:37 GMT+0000 (Coordinated Universal Time) updatedAt: Thu Jul 18 2024 13:40:04 GMT+0000 (Coordinated Universal Time) --- # MiniGpt-4 API Usage Guide ## Introduction This document will guide developers on how to use the aonweb library to call the MiniGpt-4 API, A model which generates text in response to an input image and prompt. ## Prerequisites - Node.js environment - `aonweb` library installed - Valid Aonet APPID ## Basic Usage ### 1. Import Required Modules ```js import { AI, AIOptions } from 'aonweb'; ``` ### 2. Initialize AI Instance ```js const ai_options = new AIOptions({ appId: 'your_app_id_here', dev_mode: true }); const aonweb = new AI(ai_options); ``` ### 3. Prepare Input Data Example ```js const data = { input:{ "image": "https://replicate.delivery/pbxt/IqG1MbemhULihtfr62URRZbI29XtcPsnOYASrTDQ6u5oSqv9/llama_13b.png", "top_p": 0.9, "prompt": "This llama's name is Dave. Write me a story about how Dave found his skateboard.", "num_beams": 5, "max_length": 4000, "temperature": 1.32, "max_new_tokens": 3000, "repetition_penalty": 1 } }; ``` ### 4. Call the AI Model ```js const price = 8; // Cost of the AI call try { const response = await aonweb.prediction("/predictions/ai/minigpt-4@daanelson", data, price); // Handle response console.log("MiniGpt-4 result:", response); } catch (error) { // Error handling console.error("Error generating :", error); } ``` ### Parameter Description - `image` String,Provide the image url required for model inference - `top_p` Number,Sample from the top p percent most likely tokens - `prompt` String, Prompt for mini-gpt4 regarding input image - `num_beams` Number,Number of beams for beam search decoding - `max_length` Number,Total length of prompt and output in tokens - `temperature` Number,Temperature for generating tokens, lower = more predictable results - `max_new_tokens` Number,Maximum number of new tokens to generate - `repetition_penalty` Number,Penalty for repeated words in generated text; 1 is no penalty, values greater than 1 discourage repetition, less than 1 encourage it. ### Notes - Ensure that the provided audio URL is publicly accessible and of good quality to achieve the best cloning effect. - The API may take some time to process the input and generate the result, consider implementing appropriate wait or loading states. - Handle possible errors, such as network issues, invalid input, or API limitations. - Adhere to the terms of use and privacy regulations, especially when handling voice samples of others. ### Example Response The API response will contain the URL of the generated cloned voice or other relevant information. Parse and use the response data according to the actual API documentation.