Skip to main content

Processing

Language Models​

Language models are advanced AI systems that can understand, interpret, and generate human language. These models are trained on enormous text datasets and learn patterns, word combinations, sentence structures, and even the nuances of different languages and language uses. The core of many modern language models is the transformer architecture, which uses self-attention mechanisms to determine which parts of the text are important in a given context.

In language processing, these models use statistical methods to predict the most likely next word or the most likely next sentence in a text. They can understand context over long pieces of text and thereby generate not only grammatically correct but also contextually coherent and relevant texts.

When using a language model for a chatbot or text generator, the model is given certain prompts or initial data, and based on that input, the model generates text that logically follows from the given context. The goal of these models is to produce text that seems as human-like as possible, both in terms of content and style.

Text-to-Image Models​

Text-to-image models are AI systems capable of generating visual representations from written textual descriptions, such as photos, illustrations, or other types of visual material. These models use advanced neural networks, specifically generative adversarial networks (GANs) or variations such as diffusion models.

The process begins with a text description entered by a user. The model evaluates this text and tries to understand its meaning and context. Then the model generates images that correspond to the textual description, using what it has learned during training, where it is trained on enormous datasets of text-image pairs.

During training, the model learns associations between textual descriptions and visual features. For example, if the model repeatedly sees the word combination 'a yellow sun above a blue sea' together with images illustrating this scenario, it learns to recognize and reproduce these elements in future image creations.

The result is often surprisingly accurate and detailed images that align with the entered text description. These models are becoming increasingly refined and are able to represent complex scenarios with multiple objects and abstract concepts. They are used in a wide range of applications, including artistic creations, game design, virtual reality, and more.

AI-School Unlocks Models​

It is important to understand that AI-School unlocks various AI models offered by large technology companies via an API. An API, or Application Programming Interface, is a set of rules and definitions that allow software programs to communicate with each other. It functions as a kind of 'language' understood by programs to exchange information and invoke each other's functions. AI-School does not have its own language models or text-to-image models.

We are not responsible for the results of the various models. However, we have paid attention to selecting the best and most interesting models for schools.

Processing Procedure​

The following procedure is followed to generate a response:

  • The user creates a prompt.
  • The front-end web application links this to the active chat and adds a chat message with status "Initializing".
  • A function is triggered on the AI-School servers based on an HTTP request.
  • The status of the chat message is set to "Processing".
  • When selecting a chat with documents, the server first sends a request to the Firestore vector database to select texts from documents.
  • The server then sends the request via an API connection to the selected language model.
  • Once the entire response is received, the status is set to "Completed".
  • The front-end application is refreshed.
  • In case of detected errors, the status is set to "Error" and an error message is displayed.

We do not send any personal data with each API request. However, the user may have included personal data in the prompt or in the uploaded documents.