diktn Documentation

Getting Started

Welcome to the diktn documentation! This guide will help you configure and use the Diktn text generation AI framework.


                            # Installation

                            pip install diktn

                            

                            # Quick Start

                            import diktn

                            model = diktn.load("mini-1b")

                            print(model.generate("AI is", max_new_tokens=50))

API Reference

Model Functions

diktn. load(model_name) - Loads a model from disk or HuggingFace

model. generate(prompt, max_tokens=100) - Generates text from input prompt

Examples

Text Generation

                            response = model.generate("The future of AI will", max_new_tokens=20)
                            print(response)
                            # Output: "The future of AI will bring breakthroughs in healthcare and education through accessible AI tools."
                        

Custom Models

                            model = diktn.load("custom-model", quantize=True)
                            print(model.stats())
                            # Output: {"size": "4.7GB", "layers": 24, "params": 1.4e9}
                        