Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
37 views5 pages

Generative AI Apps Overview

The document provides an overview of two generative AI applications: ChatGPT, which generates human-like text for various use cases, and DALL·E, which creates images from textual descriptions. It details their functionalities, high-level architecture, and API endpoints for generating text and images. Each application includes specific request and response formats for their respective API calls.

Uploaded by

Ankur Damor
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
37 views5 pages

Generative AI Apps Overview

The document provides an overview of two generative AI applications: ChatGPT, which generates human-like text for various use cases, and DALL·E, which creates images from textual descriptions. It details their functionalities, high-level architecture, and API endpoints for generating text and images. Each application includes specific request and response formats for their respective API calls.

Uploaded by

Ankur Damor
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Generative AI Apps Overview

1. ChatGPT (Generative Text AI)

Overview: ChatGPT is an AI language model that generates human-like text based on input.

It is commonly used for natural language conversations, content generation, code generation,

question answering, etc.

Functionality:

- Input: Text-based queries and prompts

- Output: Text-based responses

- Technology: GPT architecture (GPT-4 or other versions)

- Use Cases: Virtual assistants, customer support, education, content creation.

2. DALL·E (Generative Image AI)

Overview: DALL·E is an AI model that generates images based on textual descriptions.

It can create unique images, drawings, or paintings from detailed text prompts.

Functionality:

- Input: A text prompt describing the desired image.

- Output: Generated images or illustrations.

- Technology: A model trained on image-text pairs.

- Use Cases: Digital art creation, advertising, design inspiration, etc.

3. High-Level Architecture Design

ChatGPT Architecture:

User Input --> [ChatGPT API] --> [Preprocessing] --> [AI Model (GPT)] --> [Postprocessing] --> User
Output

| |

DB (for training data)

[External APIs/Services]

DALL·E Architecture:

User Input --> [DALL·E API] --> [Text-to-Image Model] --> Image Generation --> User Output

| |

DB (for image-text pairs)

[External Services (e.g., model fine-tuning)]

4. API Endpoint Documentation

ChatGPT API Endpoints:

1. Generate Text Completion:

- Endpoint: /v1/completions

- Method: POST

- Request Format:

"model": "gpt-4",

"prompt": "Hello, how are you today?",

"max_tokens": 50,

"temperature": 0.7

}
- Response Format:

"id": "abc123",

"object": "text_completion",

"created": 1634555067,

"model": "gpt-4",

"choices": [

"text": "I'm doing well, thank you! How can I assist you today?"

2. Stream Text:

- Endpoint: /v1/completions/stream

- Method: POST

- Request Format:

"model": "gpt-4",

"prompt": "Can you tell me about the weather?",

"stream": true

- Response Format:

"id": "xyz456",

"object": "stream",
"data": "It's sunny and warm today."

DALL·E API Endpoints:

1. Generate Image from Text:

- Endpoint: /v1/images/generations

- Method: POST

- Request Format:

"prompt": "A futuristic city with flying cars",

"n": 1,

"size": "1024x1024"

- Response Format:

"data": [

"url": "https://example.com/generated_image.png"

2. Edit Image (Inpainting):

- Endpoint: /v1/images/edits

- Method: POST
- Request Format:

"image_url": "https://example.com/input_image.png",

"mask": "The area to edit",

"prompt": "Add a sunset in the background"

- Response Format:

"data": [

"url": "https://example.com/edited_image.png"

You might also like