Skip to main content

Hyko Toolkit: Your Toolkit for Crafting AI Pipelines

The Hyko Toolkit contains a variety of AI-related functions, simplifying the integration of these functions into AI pipelines. Users can seamlessly incorporate these functions by simply dragging and dropping them from the toolkit interface into their blueprints. These functions can range from basic mathematical operations to sophisticated, fully-fledged language models, offering a wide spectrum of AI capabilities.

Functions​

These are utility functions that assist users in building their AI pipelines. Functions are categorized based on the data they manipulate or the task they perform, such as Converters, NLP or Web scraping. They enable users to perform specific actions within their pipelines, manipulate data, perform calculations...

Here's a glimpse of some of the available functions in Hyko:

  • Image: Caption, Opacity, Flip, Resize and more.
  • Converters: Pdf to String, Video to Audio and more.
  • Math: Arithmetic, Random, Range and more.
  • String: Slice, Concat, Split and more.

Models​

These are pre-trained open-source AI models that can be integrated into user's pipelines. Models are grouped into categories based on the specific task they perform, including computer vision, natural language processing, text completion and more.

Hyko toolkit offers direct access to over 130,000 models from HuggingFace, allowing users to leverage state-of-the-art models for their AI pipelines.

Here's a glimpse of some the available models in Hyko:

  • NLP: Text Classification, Question Answering, Text to Speech Generation and more.
  • Computer Vision: Image segmentation, Object Detection, Video Classification and more.
  • Regression: Linear Regression.

Inputs/Outputs​

Inputs serve as the gateway for users to provide their data to the functions, enabling them to tailor the operations to their specific needs. On the other hand, outputs serve as the means for users to observe and analyze the outcomes or results generated by these functions, ensuring that they can easily comprehend the results and make informed decisions within their AI workflows. Users can choose from a variety of input and output types, including:

  • Number: Accepts or provides numerical data including integers and decimals.
  • Text: Accepts or provides text-based input or output.
  • Boolean: Boolean input or output represents true or false values.
  • Image: Accepts or provides image data (PNG, JPEG...).
  • Audio: Accepts or provides sound or audio data (MP3, WAV, MPEG...).
  • Video: Accepts or provides video data (MP4, WEBM...)
  • PDF: Accepts or provides PDF files.

Using Hyko Toolkit Functions and Models​

To incorporate a function from Hyko Toolkit into their pipeline, users simply need to navigate to the desired category, select the function they want, and drag it into the blueprint. Once placed, they can configure the function's parameters and connect it to other nodes in the pipeline to define the workflow.

Function parameters​

Function parameters are specific inputs to functions that enable users to configure and customize their behavior. These parameters allow users to tailor the function's operation to meet specific requirements.

note

It's important to note that some function parameters are required, and not specifying them will result in an execution error.