Backprop makes it simple to use, finetune, and deploy state-of-the-art ML models.

With Backprop, you can solve a variety of tasks using pre-trained models or finetune them for your specific objectives. Here are some tasks you can tackle:

• Conversational question answering in English
• Text Classification in 100+ languages
• Image Classification
• Text Vectorisation in 50+ languages
• Image Vectorisation
• Summarisation in English
• Emotion detection in English
• Text Generation

For more specific use cases, you can adapt a task with little data, all through simple finetuning.

Getting Started

Installation

To install Backprop, simply use PyPi:

pip install backprop

Basic Task Inference

With Backprop, tasks are interfaces that allow easy access to various supported models. Here’s how you can get started:

import backprop
context = "Take a look at the examples folder to see use cases!"
qa = backprop.QA()
answer = qa("Where can I see what to build?", context)
print(answer)

Basic Finetuning and Uploading

Each task in Backprop supports finetuning that allows you to adapt models for your needs with just one line of code:

import backprop
tg = backprop.TextGeneration("t5-small")

inp = ["I really liked the service I received!", "Meh, it was not impressive."]
out = ["positive", "negative"]

tg.finetune(input_text=inp, output_text=out)
prediction = tg("I enjoyed it!")
print(prediction)

Once your model is trained, you can upload it for production use:

name = "t5-sentiment"
description = "Predicts positive and negative sentiment"
tg.upload(name=name, description=description, api_key="abc")

Why Backprop?

• No experience needed: Entrance to practical AI should be simple.
• Data is a bottleneck: Solve real-world tasks without an overwhelming amount of data.
• Curated selection: Access to the best open-source models made simple.
• Cost-effective deployment: Just call the API for most use cases.

Examples

Backprop provides various examples to get you started:

• Solve any text-based task with Finetuning: Github | Colab
• Search for images using text: Github
• Finding answers from text: Github
• More finetuning and task examples

Documentation

For extensive details on task inference and finetuning, visit our docs.

Model Hub

Check out our curated list of state-of-the-art models.

Demos

Experience zero-shot image classification with CLIP.

Troubleshooting

If you encounter any issues or have feedback, feel free to open an issue.

For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.

At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.