Xtreme1 is an innovative open-source platform designed to streamline the processes of data annotation, curation, and ontology management, particularly in the realms of computer vision and Large Language Models (LLM). This guide will take you through the installation process step by step, helping you unlock the full potential of Xtreme1’s capabilities.

Key Features of Xtreme1

• Supports data labeling for images, 3D LiDAR, and 2D/3D Sensor Fusion datasets.
• Built-in pre-labeling and interactive models for object detection, segmentation, and classification.
• Configurable Ontology Center, allowing for attribute hierarchies for model training.
• Data management and quality monitoring tools.
• Model result visualization tools for evaluation purposes.
• Reinforcement Learning from Human Feedback (RLHF) support for LLMs (beta version).

Installation Steps

Prerequisites

Before installing Xtreme1, ensure your system meets the following requirements:

• Operating Systems: Compatible with any OS using Docker Compose.
• Hardware:
  • CPU: AMD64 or ARM64
  • RAM: 2GB or higher
  • Disk Space: 10GB+ free (dependent on data size)
• Software:
  • Docker Desktop for Mac/Windows: 4.1 or newer
  • Docker Engine for Linux: 20.10 or newer
  • Docker Compose Plugin for Linux: 2.0 or newer

Step 1: Download the Package

Begin by downloading the latest release package and unzipping it. Use the following command:

wget https://github.com/xtreme1-io/xtreme1/releases/download/v0.9.1/xtreme1-v0.9.1.zip
unzip -d xtreme1-v0.9.1 xtreme1-v0.9.1.zip

Step 2: Start Services

Navigate to the directory of the unzipped package and start all services with the command:

cd xtreme1-v0.9.1
docker compose up

It may take a few minutes to initialize the database and prepare a test dataset. Once started, visit http://localhost:8190 in your browser to access Xtreme1.

Step 3: Running Built-in Models

To run the built-in models, specify a model profile:

docker compose --profile model up

Ensure that you have installed the NVIDIA CUDA Driver and the NVIDIA Container Toolkit if you’re using built-in models on a Linux server.

Understanding Packages with an Analogy

Think of setting up Xtreme1 similar to preparing a big meal for a special dinner. Each step represents preparing different components of the meal:

• Downloading the Package: This is like buying all your ingredients—get everything fresh and ready before cooking!
• Starting Services: This is akin to preheating your oven and starting your stove—the cooking itself won’t happen until everything is ready!
• Running Built-in Models: This step is like actually cooking the meal. You need to follow the right processes so that the meal is both delicious and successful!

Troubleshooting Tips

• Ensure that you have Docker installed correctly. Refer to the Docker Desktop documentation for installation issues.
• If you face problems with model compatibility on an ARM CPU, create a Docker Compose override file as described in the README.
• If encountering package version issues, you may want to clear the builder cache using the command: docker builder prune.

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

Conclusion

Xtreme1 offers a powerful platform for multimodal training data, allowing users to maximize their efficiency in data annotation and management. With its user-friendly setup and feature-rich environment, it’s an essential tool for both new and experienced AI developers. 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.